Wood inspection with non-supervised clustering
- 格式:pdf
- 大小:1.57 MB
- 文档页数:13
European Medicines Agency Evaluation of Medicines for Human Use7 Westferry Circus, Canary Wharf, London, E14 4HB, UKTel. (44-20) 74 18 84 00 Fax (44-20) 74 18 86 13 London, 28 June 2006 CPMP/SWP/5199/02EMEA/CHMP/QWP/251344/2006 COMMITTEE FOR MEDICINAL PRODUCTS FOR HUMAN USE(CHMP)GUIDELINE ON THE LIMITS OF GENOTOXIC IMPURITIES DISCUSSION IN THE SAFETY WORKING PARTY June 2002 – October2002 TRANSMISSION TO CPMP December 2002 RELEASE FOR CONSULTATION December 2002 DEADLINE FOR COMMENTS March 2003DISCUSSION IN THE SAFETY WORKING PARTY AND QUALITY WORKING PARTYJune 2003 - February 2004TRANSMISSION TO CPMP March 2004 RE-RELEASE FOR CONSULTATION June 2004 DEADLINE FOR COMMENTS December 2004DISCUSSION IN THE SAFETY WORKING PARTY AND QUALITY WORKING PARTY February 2005 -May2006ADOPTION BY CHMP 28 June 2006 DATE FOR COMING INTO EFFECT 01 January 2007KEYWORDS Impurities; Genotoxicity; Threshold of toxicological concern (TTC);Structure activity relationship (SAR)GUIDELINE ON THE LIMITS OF GENOTOXIC IMPURITIESTABLE OF CONTENTSEXECUTIVE SUMMARY (3)1.INTRODUCTION (3)2.SCOPE (3)3.LEGAL BASIS (3)4.TOXICOLOGICAL BACKGROUND (4)5.RECOMMENDATIONS (4)5.1Genotoxic Compounds With Sufficient Evidence for a Threshold-Related Mechanism (4)5.2Genotoxic Compounds Without Sufficient Evidence for a Threshold-Related Mechanism (5)5.2.1Pharmaceutical Assessment (5)5.2.2Toxicological Assessment (5)5.2.3Application of a Threshold of Toxicological Concern (5)5.3Decision Tree for Assessment of Acceptability of Genotoxic Impurities (7)REFERENCES (8)EXECUTIVE SUMMARYThe toxicological assessment of genotoxic impurities and the determination of acceptable limits for such impurities in active substances is a difficult issue and not addressed in sufficient detail in the existing ICH Q3X guidances. The data set usually available for genotoxic impurities is quite variable and is the main factor that dictates the process used for the assessment of acceptable limits. In the absence of data usually needed for the application of one of the established risk assessment methods, i.e. data from carcinogenicity long-term studies or data providing evidence for a threshold mechanism of genotoxicity, implementation of a generally applicable approach as defined by the Threshold of Toxicological Concern (TTC) is proposed. A TTC value of 1.5 µg/day intake of a genotoxic impurity is considered to be associated with an acceptable risk (excess cancer risk of <1 in 100,000 over a lifetime) for most pharmaceuticals. From this threshold value, a permitted level in the active substance can be calculated based on the expected daily dose. Higher limits may be justified under certain conditions such as short-term exposure periods.1. INTRODUCTIONA general concept of qualification of impurities is described in the guidelines for active substances (Q3A, Impurities in New Active Substances) or medicinal products (Q3B, Impurities in New Medicinal Products), whereby qualification is defined as the process of acquiring and evaluating data that establishes the biological safety of an individual impurity or a given impurity profile at the level(s) specified. In the case of impurities with a genotoxic potential, determination of acceptable dose levels is generally considered as a particularly critical issue, which is not specifically covered by the existing guidelines.2. SCOPEThis Guideline describes a general framework and practical approaches on how to deal with genotoxic impurities in new active substances. It also relates to new applications for existing active substances, where assessment of the route of synthesis, process control and impurity profile does not provide reasonable assurance that no new or higher levels of genotoxic impurities are introduced as compared to products currently authorised in the EU containing the same active substance. The same also applies to variations to existing Marketing Authorisations pertaining to the synthesis. The guideline does, however, not need to be applied retrospectively to authorised products unless there is a specific cause for concern.In the current context the classification of a compound (impurity) as genotoxic in general means that there are positive findings in established in vitro or in vivo genotoxicity tests with the main focus on DNA reactive substances that have a potential for direct DNA damage. Isolated in vitro findings may be assessed for in vivo relevance in adequate follow-up testing. In the absence of such information in vitro genotoxicants are usually considered as presumptive in vivo mutagens and carcinogens.3. LEGAL BASISThis guideline has to be read in conjunction with Directive 2001/83/EC (as amended) and all relevant CHMP Guidance documents with special emphasis on:Impurities Testing Guideline: Impurities in New Drug Substances (CPMP/ICH/2737/99, ICHQ3A(R)) Note for Guidance on Impurities in New Drug Products (CPMP/ICH/2738/99, ICHQ3B (R))Note for Guidance on Impurities: Residual Solvents (CPMP/ICH/283/95)Note for Guidance on Genotoxicity: Guidance on Specific Aspects of Regulatory Genotoxicity Tests for Pharmaceuticals (CPMP/ICH/141/95, ICHS2A)Note for Guidance on Genotoxicity: A Standard Battery for Genotoxicity Testing of Pharmaceuticals (CPMP/ICH/174/95, ICHS2B)4. TOXICOLOGICAL BACKGROUNDAccording to current regulatory practice it is assumed that (in vivo) genotoxic compounds have the potential to damage DNA at any level of exposure and that such damage may lead/contribute to tumour development. Thus for genotoxic carcinogens it is prudent to assume that there is no discernible threshold and that any level of exposure carries a risk.However, the existence of mechanisms leading to biologically meaningful threshold effects is increasingly acknowledged also for genotoxic events. This holds true in particular for compounds interacting with non-DNA targets and also for potential mutagens, which are rapidly detoxified before coming into contact with critical targets. The regulatory approach to such chemicals can be based on the identification of a critical no-observed-effect level (NOEL) and use of uncertainty factors.Even for compounds which are able to react with the DNA molecule, extrapolation in a linear manner from effects in high-dose studies to very low level (human) exposure may not be justified due to several protective mechanisms operating effectively at low doses. However, at present it is extremely difficult to experimentally prove the existence of threshold for the genotoxicity of a given mutagen. Thus, in the absence of appropriate evidence supporting the existence of a threshold for a genotoxic compound making it difficult to define a safe dose it is necessary to adopt a concept of a level of exposure that carries an acceptable risk.5. RECOMMENDATIONSAs stated in the Q3A guideline, actual and potential impurities most likely to arise during synthesis, purification and storage of the new drug substance should be identified, based on a sound scientific appraisal of the chemical reactions involved in the synthesis, impurities associated with raw materials that could contribute to the impurity profile of the new drug substance, and possible degradation products. This discussion can be limited to those impurities that might reasonably be expected based on knowledge of the chemical reactions and conditions involved. Guided by existing genotoxicity data or the presence of structural alerts, potential genotoxic impurities should be identified. When a potential impurity contains structural alerts, additional genotoxicity testing of the impurity, typically in a bacterial reverse mutation assay, should be considered (Dobo et al. 2006, Müller et al. 2006). While according to the Q3A guideline such studies can usually be conducted on the drug substance containing the impurity to be controlled, studies using isolated impurities are much more appropriate for this purpose and highly recommended.For determination of acceptable levels of exposure to genotoxic carcinogens considerations of possible mechanisms of action and of the dose-response relationship are important components. Based on the above considerations genotoxic impurities may be distinguished into the following two classes:- Genotoxic compounds with sufficient (experimental) evidence for a threshold-related mechanism - Genotoxic compounds without sufficient (experimental) evidence for a threshold-related mechanism5.1 Genotoxic Compounds With Sufficient Evidence for a Threshold-Related Mechanism Examples of mechanisms of genotoxicity that may be demonstrated to lead to non-linear or thresholded dose-response relationships include interaction with the spindle apparatus of cell division leading to aneuploidy, topoisomerase inhibition, inhibition of DNA synthesis, overloading of defence mechanisms, metabolic overload and physiological perturbations (e.g. induction of erythropoeisis, hyper- or hypothermia).For (classes of) compounds with clear evidence for a thresholded genotoxicity, exposure levels which are without appreciable risk of genotoxicity can be established according to the procedure as outlined for class 2 solvents in the Q3C Note for Guidance on Impurities: Residual Solvents. This approach calculates a “Permitted Daily Exposure” (PDE), which is derived from the NOEL, or the lowest-observed effect level (LOEL) in the most relevant (animal) study using “uncertainty factors” (UF).5.2 Genotoxic Compounds Without Sufficient Evidence for a Threshold-Related Mechanism The assessment of acceptability of genotoxic impurities for which no threshold mechanisms are identified should include both pharmaceutical and toxicological evaluations. In general, pharmaceutical measurements should be guided by a policy of controlling levels to “as low as reasonably practicable” (ALARP principle), where avoiding is not possible. Levels considered being consistent with the ALARP principle following pharmaceutical assessment should be assessed for acceptability from a toxicological point of view (see decision tree & following sections).5.2.1 Pharmaceutical AssessmentA specific discussion – as part of the overall discussion on impurities (see Q3A(R)) – should be provided in the application with regard to impurities with potential genotoxicity.A rationale of the proposed formulation/manufacturing strategy should be provided based on available formulation options and technologies. The applicant should highlight, within the chemical process and impurity profile of active substance, all chemical substances, used as reagents or present as intermediates, or side-products, known as genotoxic and/or carcinogenic (e.g. alkylating agents). More generally, reacting substances and substances which show “alerting structure” in terms of genotoxicity which are not shared with the active substance should be considered (see e.g. Dobo et al. 2006). Potential alternatives which do not lead to genotoxic residues in the final product, should be used if available.A justification needs to be provided that no viable alternative exists, including alternative routes of synthesis or formulations, different starting materials. This might for instance include cases where the structure, which is responsible for the genotoxic and/or carcinogenic potential is equivalent to that needed in chemical synthesis (e.g. alkylation reactions).If a genotoxic impurity is considered to be unavoidable in a drug substance, technical efforts (e.g. purification steps) should be undertaken to reduce the content of the genotoxic residues in the final product in compliance with safety needs or to a level as low as reasonably practicable (see safety assessment). Data on chemical stability of reactive intermediates, reactants, and other components should be included in this assessment.Detection and/or quantification of these residues should be done by state-of-the-art analytical techniques.5.2.2 Toxicological AssessmentThe impossibility of defining a safe exposure level (zero risk concept) for genotoxic carcinogens without a threshold and the realization that complete elimination of genotoxic impurities from drug substances is often unachievable, requires implementation of a concept of an acceptable risk level, i.e. an estimate of daily human exposure at and below which there is a negligible risk to human health. Procedures for the derivation of acceptable risk levels are considered in the Appendix 3 of the Q3C Note for Guidance on Impurities: Residual Solvents for Class 1 solvents. However, these approaches require availability of adequate data from long-term carcinogenicity studies.In most cases of toxicological assessment of genotoxic impurities only limited data from in vitro studies with the impurity (e.g. Ames test, chromosomal aberration test) are available and thus established approaches to determine acceptable intake levels cannot be applied. Calculation of “safety multiples” from in vitro data (e.g. Ames test) are considered inappropriate for justification of acceptable limits. Moreover, negative carcinogenicity and genotoxicity data with the drug substance containing the impurity at low ppm levels do not provide sufficient assurance for setting acceptable limits for the impurity due to the lack of sensitivity of this testing approach. Even potent mutagens and carcinogens are most likely to remain undetected when tested as part of the drug substance, i.e. at very low exposure levels. A pragmatic approach is therefore needed which recognises that the presence of very low levels of genotoxic impurities is not associated with an unacceptable risk.5.2.3 Application of a Threshold of Toxicological ConcernA threshold of toxicological concern (TTC) has been developed to define a common exposure level for any unstudied chemical that will not pose a risk of significant carcinogenicity or other toxic effects(Munro et al. 1999, Kroes and Kozianowski 2002). This TTC value was estimated to be 1.5 µg/person/day. The TTC, originally developed as a “threshold of regulation” at the FDA for food-contact materials (Rulis 1989, FDA 1995) was established based on the analysis of 343 carcinogens from a carcinogenic potency database (Gold et al. 1984) and was repeatedly confirmed by evaluations expanding the database to more than 700 carcinogens (Munro 1990, Cheeseman et al. 1999, Kroes et al. 2004). The probability distribution of carcinogenic potencies has been used to derive an estimate of a daily exposure level (µg/person) of most carcinogens which would give rise to less than a one in a million (1 x 10-6) upper bound lifetime risk of cancer (“virtually safe dose”). Further analysis of subsets of high potency carcinogens led to the suggestion of a 10-fold lower TTC (0.15 µg/day) for chemicals with structural alerts that raise concern for potential genotoxicity (Kroes et al. 2004). However, for application of a TTC in the assessment of acceptable limits of genotoxic impurities in drug substances a value of 1.5 µg/day, corresponding to a 10-5 lifetime risk of cancer can be justified as for pharmaceuticals a benefit exists. It should be recognized in this context that the methods on which the TTC value is based, are generally considered very conservative since they involved a simple linear extrapolation from the dose giving a 50% tumour incidence (TD50) to a 1 in 106 incidence, using TD50 data for the most sensitive species and most sensitive site (several “worst case” assumptions) (Munro et al. 1999).Some structural groups were identified to be of such high potency that intakes even below the TTC would be associated with a high probability of a significant carcinogenic risk (Cheeseman et al. 1999, Kroes et al. 2004). This group of high potency genotoxic carcinogens comprises aflatoxin-like-, N-nitroso-, and azoxy-compounds that have to be excluded from the TTC approach. Risk assessment of members of such groups requires compound-specific toxicity data.There may be reasons to deviate from the TTC value based on the profile of genotoxicity results. Positive result from in vitro studies only may allow to exempt an impurity from limitation at TTC level if lack of in vivo relevance of the findings is convincingly demonstrated based on a weight-of-evidence approach (see ICH S2 guidelines). This approach will usually need negative results with the impurity from some additional in vitro and/or appropriate in vivo testing.A TTC value higher than 1.5 µg/day may be acceptable under certain conditions, e.g. short-term exposure, for treatment of a life-threatening condition, when life expectancy is less than 5 years, or where the impurity is a known substance and human exposure will be much greater from other sources (e.g. food). Genotoxic impurities that are also significant metabolites may be assessed based on the acceptability of the metabolites.The concentration limits in ppm of genotoxic impurity in drug substance derived from the TTC can be calculated based on the expected daily dose to the patient using equation (1).(1) Concentration limit (ppm) = TTC [µg/day]dose (g/day]The TTC concept should not be applied to carcinogens where adequate toxicity data (long-term studies) are available and allow for a compound-specific risk assessment.It has to be emphasized that the TTC is a pragmatic risk management tool using a probabilistic methodology, i.e. there is a high probability that a 10-5 lifetime cancer risk will not be exceeded if the daily intake of a genotoxic impurity with unknown carcinogenic potential/potency is below the TTC value. The TTC concept should not be interpreted as providing absolute certainty of no risk.5.3 Decision Tree for Assessment of Acceptability of Genotoxic Impurities(shaded boxes = pharmaceutical assessment, white boxes = toxicological assessment)1) Impurities with structural relationship to high potency carcinogens (see text) are to beexcluded from the TTC approach2) If carcinogenicity data available: Does intake exceed calculated 10-5 cancer lifetime risk?3) Case-by-case assessment should include duration of treatment, indication, patient population etc (see text) *) Abbreviations: NOEL/UF – No Observed Effect Level/Uncertainty Factor,PDE – Permitted Daily Exposure, TTC – Threshold of Toxicological ConcernREFERENCESCheeseman M.A., Machuga E.J., Bailey A.B., A tiered approach to threshold of regulation, Food Chem Toxicol 37, 387-412, 1999Dobo K.L., Greene N., Cyr M.O., Caron S., Ku W.W., The application of structure-based assessment to support safety and chemistry diligence to manage genotoxic impurities in active pharmaceutical ingredients during drug development, Reg Tox Pharm 44, 282-293, 2006.Gold L.S., Sawyer C.B., Magaw R., Backman G.M., de Veciana M., Levinson R., Hooper N.K., Havender W.R., Bernstein L., Peto R., Pike M.C., Ames B.N., A carcinogenic potency database of the standardized results of animal bioassays, Environ Health Perspect 58, 9-319, 1984.Kroes R., Renwick A.G., Cheeseman M., Kleiner J., Mangelsdorf I., Piersma A., Schilter B., Schlatter J., van Schothorst F., Vos J.G., Würtzen G., Structure-based threshold of toxicological concern (TTC): guidance for application to substances present at low levels in the diet, Food Chem Toxicol 42, 65-83, 2004.Kroes R., Kozianowski G., Threshold of toxicological concern (TTC) in food safety assessment, Toxicol Letters 127, 43-46, 2002.Müller L., Mauthe R.J., Riley C.M., Andino M.M., De Antonis D., Beels C., DeGeorge J., De Knaep A.G.M., Ellison D., Fagerland J.A., Frank R., Fritschel B., Galloway S., Harpur E., Humfrey C.D.N., Jacks A.S.J., Jagota N., Mackinnon J., Mohan G., Ness D.K., O’Donovan M.R., Smith M.D., Vudathala G., Yotti L., A rationale for determining, testing, and controlling specific impurities in pharmaceuticals that possess potential for genotoxicity, Reg Tox Pharm 44, 198-211, 2006Munro I.C., Safety assessment procedures for indirect food additives: an overview. Report of a workshop. Reg Tox Pharm 12, 2-12, 1990.Munro I.C., Kennepohl E., Kroes R., A procedure for the safety evaluation of flavouring substances, Food Chem Toxicol 37, 207-232, 1999.Rulis A.M., Establishing a threshold of regulation. In Risk Assessment in Setting National Priorities (J.J. Bonin and D.E. Stevenson, Eds.) Plenum, New York, 271-278, 1989.U.S. Food and Drug Administration (FDA), Food additives: Threshold of regulation for substances used in food-contact articles (final rule), Fed. Regist. 60, 36582-36596, 1995.。
施工工艺说明,中英文翻译目目录录目目录录SECTION A 3SECTION B 11SECTION C 21SECTION D 36SECTION E 49SECTION F58SECTION G 71SECTION H 79SECTION I942SECTION A第一章第一章第一章第一章CARPENTER AND JOINER木匠和工匠木匠和工匠木匠和工匠木匠和工匠10 SCOPE OF WORK工程范围The works in this specification shall include but not necessarilybe limited tothe supply and installation of the following 此技术标准的工程范围包括但不限于如下产品的提供和安装 Timber Doors and Frames Installation木门和门框的安装Hardware Ironmongery and Door Furniture Installation 五金件铁制品和家具门安装Sundry Joinery cabinets counters Installation 各种木制产品橱柜和柜台安装Timber Flooring and Decking Supply and Installation 木制地板和装饰品供应和安装Built-in Furniture and Cabinetry Supply and Installation内置家具和细木家具的供应和安装20 CARPENTRY JOINERY GENERALLY木匠和工匠概要21 Timber shall be first quality of the specifies specified straight grainedproperly seasoned free from sap and free from shakes large and looseknots木制品首先要符合特殊技术标准文理清晰做工精细无腐蚀抗震无大的树眼22 Timber for both carpentry and joinery work shall be tanalised localhardwoodunless otherwise stated除非特殊说明木匠和工匠制造的木制品将为用防腐剂处理过的当地硬木23 Scantlings shall be in long lengths sawn square to the required sizes freefrom twist or curl with a trade allowance of 3mm- for a sawn timber beingpermitted A imum of 3mm additional allowance for dressing will bepermitted从长木条中抽取一小块按照原标准的正方形随意扭转弯折的误差在3毫米内最终的完成面的的误差范围在3毫米内24 All scantlings for studs noggins plates etc shall be gauged for uniformityof thickness所有少量的钉子五金配件金属板等都将统一厚度标准25 Samples of all timbers to be used shall be submitted to the ProjectManager in triplicate for approval and once approved the Contractor shallorder the timber as soon as possible after the Contract is let to allow time formilling the quantity and sizes required and for adequate drying of same所有提交给项目公司的木制品的样品都应得到批准经过批准的承包商应立即按照合同规定的相同的烘干尺寸和型号订购木料326 All timbers to be used shall be seasoned so that the moisture content isbetween 12 and 15 The Contractor shall submit written proof ofmoisture content from an approved laboratory or arrange for a site test to besupervised by the Project Manager whenever called upon to do so所有要使用的木料都应是加工过的湿度在12到15之间承包商将提交实验室出具的湿度书面证明或是安排项目经理人随时到现场进行测试27 The whole of the Carpentry and Joinery works throughout shall beexecuted in the best and most workmanlike manner All necessarytemplates lining blocks stops etc shall be provided and fixed and alltrimmings grooving tonguing mitering etc incidental to the work shall bedone even though not specifically mentioned所有木匠和连接工程都将保证做到技术娴熟提供并加工所有必需的模板内置木板木楔等从事所有和工程有关的即使没有提到的过程或方法28 All timber exposed to view shall be wrot所有木材外表面应感观良好30 SCHEDULE OF TIMBERS木材部分com Unless otherwise described all timber for Carpenters and Joinersworkshall be in accordance with the finishesschedule as specified in thefinishes Schedule All timber framework behindceilings paneling etcshall be first quality local hardwood with aspecific gravity of 075 45lbscu ft strength to be group "B"除非特殊提出所有木匠和工匠使用的木材将符合装饰表格内对装饰的特殊说明天花板嵌板等所有木材都采用当地合格的硬木其容重为075 45lbscu ft 强度为"B"40 PRESSURE IMPREGNATION OF TIMBER真空注压压力板生产41 Timber described as "pressure-treated" shall be impregnated undervacuum and pressure withan approved local proprietary wood preservation tothe following nett dry salt retention41通过真空注压生产的板材应符合当地批准的木材干燥保存的标准标准如下Use Tanalith "C"Structural timber and 035 lb per cu ft joinery for use in theinterior of buildings above ground Exterior such as weather 050lb per cuftboard door and window frames and sill above ground在建筑物内部使用防腐剂"C"结构的木材和035 lb per cu 的木板地面以上外部如050 lb per cuft 的木板门窗框窗台地上部分42 On delivery to the site of any pressure-treated timber inspectioncertificates shall be produced for scrutiny by the Project Manager 运至施工现场的任何压好的板材检验证书将由项目经理详细审查43 All timbers shall be cut to final dimensions before impregnation Any4timber which is rip sawnor deeply wrot after impregnation shall be impregnated again as foruntreated timber to the entire satisfaction of the Project Manager所有木材在经过注入式处理前都将切割成最终尺寸所有有裂缝暴露在外或深度受损的木材将重新注入处理以使项目经理满意44 Ends of pressure-treated timber which are cut shall be sealed withapproved proprietary preservative applied strictly in accordance with themanufacturers instructions最后经过蒸压处理的木材将严格根据制造商说明进行相应严格的防腐处理并贴上合格标签45 Acclimatisation of timber to site conditions shall require that any solidtimber stock eg flooring millwork etc for internal use be delivered tosite and maintained under the same climactic conditions temperature andhumidity as would be the case for the finished project This period shouldnot be less than 10 days符合条件的木材需要采购坚实的木材作原料如地板磨坊等在相同的环境下温度和湿度运输到施工现场以便长期使用完成的项目也同样按此要求进行这段期限将不少于10天46 Plywood shall be of the grades specified in BS 1455 or local equivalentwhich in the opinion ofthe Project Manager is appropriate to the work as follows胶合板将符合英标BS 1455 等级或是项目经理人认为的适合施工的当地标准详情如下com Grade 1 - For use in the natural state the veneer shall be free fromknotsworm and beetle holes splits glue stains or other defects等级1- 用于自然状态胶合板将不带结点虫洞裂缝胶水污点或其他的缺陷com Grade 2 - For use where subsequent painting or similar treatment isrequired The veneer on the face side may have a few sound knotsoccasional minor discolouration or stain and small inlay repairs等级2- 用于持续喷漆或相似的需要胶合板的表面可有些死结小污点或瑕疵和小补丁com Grade 3 - For use where the plywood is not normally visible The veneermay have defects other than those specified above provided that theserviceability is not affected等级3- 用于合板不可见的部分如果不影响实用性胶合板可有除上述特殊强调的缺点47 Unless otherwise specified plywood shall be Grade 2 with Type 1adhesive the veneer being rotary cut Meranti of standard thicknesses asdetailed on the respective drawings除非特殊强调合板将使用粘合剂型号1等级2 经过旋削的胶合板将符合图纸中的厚度标准50 PLYWOOD胶合板Each sheet of plywood delivered to the site shall bear the followingmarkings提供给现场的胶合板应该有如下标记5a Manufacturers name or identification制造商的名字或认证b BS No 1455 or local equivalent符合英国标准BS1455或当地同等标准c Wbp weather and boil proof or where used for weather resistant locationsMR moisture and weather resistant or markings from local StandardsInstituteWbp 天气和防温性或天气场所 MR 湿度和抗环境的能力或当地标准研究所的标记d Grade of finish装饰等级e Thickness厚度60 LAMINATED PLASTIC SHEETING饰面薄板技术要求Decorative laminated plastic sheeting shall generally comply with therequirements of BS3794 or local equivalent and shall be an approvedmelamine surface sheeting 2mm in thickness of "cigarette-proof" gradeSamples of the type the contractor proposes using indicating also thetexture and pattern shall be submitted to the Project Manager for selectionand approval before use装饰的薄板状的造型防护墙将符合BS3794的标准或当地同等标准并且是批准的三聚氰胺表面厚度2毫米防烟等级在使用批准前承包商建议样品使用的标识质地型号将提交给项目经理人进行筛选和批准70 NAILS钉子Nails shall be wire or cut type of mild or black rolled steel complying therequirements of BS1202 or local equivalent and unless otherwise specifiedto be "bright" finish Galvanized or cadmium coated nails shall comply withthe requirements of BS1202 or local equivalent根据BS1202的规定或当地规定钉子将是金属丝制的或切割形状或者黑色卷钢的在精装的技术标准除外镀锌的或镉制的钉子要符合BS1202的规定或当地政府的同等规定80 POWER DRIVEN FIXINGS安装工具设备Power driven fixings shall be approved hardened steel pins driven with anapproved power charged tool "Ramset" or similar equivalent The Contractorwill be responsible for compliance with all local regulations regarding thepurchase storage and safety measures to be undertaken in the use of suchsystems and shall follow in detail the manufacturers recommendationsregarding the type and size of fastener suitable for the work被批准的安装工具是"Ramset"品牌的工具或类似品牌承包商根据当地规定购买存储安全检验负责并且遵循制造商对适合此项工程使用的工具的尺寸或型号的推荐标准90 BOLTS PLUGS SCREWS6螺钉插拴和螺丝91 BOLTS螺钉com Bolts shall be mild steels bolts with nuts and washers螺钉将是带螺母和垫圈的中型的钢螺钉com Expansions bolts shall be of approved manufacture膨胀螺钉必须是得到政府批准com All bolts used externally to be complete with nuts and washers and shallbe galvanized所有长期使用的螺钉将带有螺母和垫圈并且将是镀锌的com Bolts connecting timbers loaded parallel to the grain shall be spaced tominimum of eight times the diameter of the bolt from the end of the timberBolts connecting timber loaded perpendicular to the grain shall be spaceda minimum of four times the diameter of the bolt from the load edge Nobolt shall be closer than twice the diameter of the bolt to any edge oftimber Minimum spacing shall be two and one half times the diameter ofthe bolt由螺钉平行连接的木边应为被连接部份为木边的八倍由螺钉垂直连接的木边应为被连接部分的四倍螺钉至木边的距离不小于两倍钉子直径最小应为25倍钉子直径92 PLUGS插头Plugs shall be patent plugs of approved manufacture and shall be of theappropriate recommended size for the location required and shall be ofthe pre-formed nylon type插头将是由制造商批准的专利插头并且将是符合当地标准的被推荐的型号并将是预制尼龙的93 SCREWS螺丝com Wood screws shall comply with the requirements of BS1210 or equal andshall be sufficient size and adequate for the purpose for whichthey arerequired Unless otherwise specified brass screws shall be used for bothCarpenters and Joiners work木制螺丝将符合BS1210的规定或相关规定其型号将符合实际情况除非特殊说明铜螺丝将用于木匠和工匠的工程com Screws shall be satin chrome where exposed螺丝的外露部分将是光滑的铬com Exposed screws in locations accessible to the public shall be "Phillip"head screws当地符合标准的进入市场的螺丝为菲利浦螺丝com Screws which are intended to be removed periodically shall be fitted withsatin chrome plate cups定期可移除的螺钉应有镀铬的螺钉垫com Drill guide holes for all acrews and countersunk for heads of internalscrews other than the screws in cups Screws which become distorted orwhose head is damaged shall be removed and replaced所有螺丝和埋头钉的钻导孔都是使用的钉头形成的扭曲的或顶头损坏的螺丝将被更换和替代7com Metal thread self-tapping screws shall be of approved manufacture金属线螺钉将由制造商批准94 COACH SCREWS十字螺丝Shall be first quality mild steel of approved manufacture with hexagonalheads Each screwshall be fitted with a soft aluminium washer 3mm thick是经过国家许可的一级质量的低碳钢六角形螺丝每个螺丝锚固时将使用3毫米厚的铝制垫圈100 ADHESIVES粘合剂101 Adhesives used for constructional work and joinery in timber shall be coldsetting casein glue complying with the requirements of BS1444 or coldsetting synthetic resin adhesive moisture resistant gap filling typecomplying with the requirement of BS 1204 wep or local equivalent Theuse of animal glues is not permitted用于结构和木制品的粘结剂将是酪素胶的且符合BS1444的标准或合成树脂符合BS 1204的要求的填充剂填充缝隙动物胶禁止使用102 Adhesives for plywood shall be one of the following types胶合板的粘合剂将符合如下型号之一Type 1 - Pheno formaldehyde resin adhesive classified as weather and boilproof Wbp inBS1203 or equal型号1-在BS1203类别为耐候胶系列的甲醛树脂粘合剂或同类产品Type 2 - Urea formaldehyde resin adhesive classified as moisture and weatherresistant MR型号2- 类别为抗湿的尿素甲醛树脂粘合剂103 Adhesives for bonding plastic sheeting to timber shall be of anapprovedsynthetic resin adhesive used strictly in accordance with themanufacturers instructions Plastic sheeting must not be applied to timberat a temperature less than 18 degrees C or with a moisture content ofmore than 15木材塑料粘合剂应为符合国家标准的合成树脂粘合剂应用于木材背后的塑胶护板粘合温度应低于18度且湿度不大于15110 DEFECTS IN CARPENTERS AND JOINERS WORK木匠和工匠工程的过失Should joints in Carpenters or Joiners work open or other defects arisewithin the period stated for maintenance in the Contract and the causethereof be deemed by the ProjectManager to be due to unseasoned timber or faulty or bad workmanshipsuch defective joinery shall be taken down refitted redecorated orreplaced if necessary and any work disturbed around shall be made goodall at the Contractors expense一旦木匠和木质的工程开始以后或在合同履行期间出现过失而且过失是由项目经理人使用不合格的木材或者工艺立即停止过失在必要得部分进行重新装修装饰替换挽回工程过失承包商将负担一切费用120 DOORS AND DOOR FRAMES8门和门框121 Flush doors shall be hollow core framed up in selected first qualityhardwood and covered with 6mm plywood as described bonded underpressure on both faces The direction of the grain of face veneer is to bevertical or as otherwise indicated on the drawings齐边拉门应为质量上成的硬木和双面捆绑的6毫米的胶合板薄板的表面应是垂直的或者在图纸上标识出122 The finished surfaces of plywood facing shall not be subject to胶合板的装饰面不应为a Lifting at edges边缘起翘b Blistering起鼓的c Sinking or raising of the surface due to defects of the base material由于基底材料的问题致使局部凹陷鼓起123 All doors shall be square and of a smooth finish with all framing properlymortised and tenoned together所有的门应是方正平滑正确榫接成的124 All doors shall have a hardwood lipping mitred around all edges所有的门应是边实木门边125 Hollow core door 44mm thick overall shall be framed up in first qualitytanalized hardwood skeleton framing with 100 x 32mm stiles top andthree intermediate rails and 150 x 32mm bottom rail faced both sides with6mm plywood with veneer finish as per finishes specifications andwith 44x 6mm edge strips all round or as detailed on drawings44米厚全部的空心门应使用质量上成的用防腐剂处理过的100 x 32 mm厚钢骨构架硬木根据装饰标准的3个中钉两面150 x 32mm的低钉 44x 6mm 的边沿衬条或按图纸的详细说明126 Doors to all kitchen and bathroom shall be of similar construction to theabove but faced both sides with 6mm marine plywood with finish as perfinishes specification and with 44 x 6mm edge strips all round所有厨房和浴室的门应符合上面提到的相似结构但按照装修技术标准两面为6 mm的胶合板 44 x 6mm 的边缘衬条127 Solid core door 44mm thick overall shall be framed up in first qualitytanlized hardwoodboarding 150 x 32mm tongued and grooved togetherand covered on both sides with 6mm plywood in specified veneer finishand 44 x 6mm edge strips all round The whole door is to be bolted with 3numbers 12mm diameter mild steel tie rods and tightened with nuts andwashers and all holes plugged and pelleted over44 mm厚全部的实心门应是特殊薄板装饰的质量上成的6 mm厚的胶合板装框和 44 x 6mm 的边缘衬条整个门是 36 mm直径的软钢把手并用螺母和垫圈固定所有的门孔都硬塞紧128 Door frames and linings shall be first quality tanlized dressed hardwood oras indicated on finishes specifications and to the profiles shown on thedetailed drawings are to be plugged screwed and pelleted to hardwoodgrounds Where required fix with approved pattern mildsteel holdfasts to jambs of brick walls with dowels at foot of frame门框和内层应选用质量上成的用防腐剂处理过的硬木或在技术标准或图纸9剖面中特殊指明的材料并在胶合板上塞紧用螺丝拧紧在需要的地方在门框底部用被核准的软钢固定砖头并用木钉拧紧129 Door jambs and heads shall be properly framed and wedged togetherwith horns removedbefore plugging to masonry门侧面和顶部应加上合适的外框在用木塞填堵之前固定四边END OF SECTION AA部分的结尾部分的结尾部分的结尾部分的结尾10SECTION B第二章第二章第二章第二章SUSPENDED CEILING AND CEILING FINISHES吊顶和天花板的装修SCOPE OF WORK工程范围The works in this specification shall include but not necessarily be limited to thesupply and installation of the following技术标准中此项工程将包括但不限于如下的供应和安装Supply and installation of all hard Cement Board Cemboard ceilings所有水泥压力 Cemboard 天花板的供应和安装Supply and installation of all Plasterboard Ceilings所有石膏天花板的供应和安装Supply and installation of mineral fiber ceilings矿物质纤维天花板的供应和安装Co-ordination with all M E trades in relation to the installation of all M Eworks prior toceiling works commencing在开展与天花板有关的工程前就机电安装问题与机电供应商进行协调Co-ordination with all M E trades in relation to the locations of all outletsequipment andaccess openings and other installations which will affect ceiling framing和机电供应商就插座设备检修孔以及其他影响天花板装饰的安装进行协调Statutory Provisions Standards and Codes of Practice法定条款标准和实际编号All work and materials shall except where otherwise indicated on the drawings ordirected in this Specification comply with the following除了在图纸上或技术标准中特殊标识外所有的工程和材料将符合如下标准AS 1045 Measurement of Absorption Co efficient in a Reverberation RoomAS 1045混响室的音响吸收标准AS 1214 Hot-Dip Galvanized Coatings on Threaded FastenersAS 1214热镀锌表面的罗纹紧固件AS 1397 Steel Sheet and Strip –Hot-Dipped Zinc-Coated or AluminiumZinc CoatedAS 1397钢板和钢条热镀锌或铝锌的表层AS 1789 Electroplated Coating – Zinc on Iron or SteelAS 1789 electroplated coating电镀锌铁或钢的AS 2107 Acoustics –Recommended Design Sound Levels and Reverberation Timesfor BuildingInteriors11AS 2107音响–被推荐的音响设计标准和室内建筑的回响次数AS 2499 Acoustics –AS 2499 Method for Laboratory Measurement of AirborneSound Attenvation of CeilingsAS 2499音响–AS 2499天花板空气载声的实验室测量方法AS 2588 Gypsum PlasterboardAS 2588石膏板AS 2590 Glass Fibre Reinforced Gypsum PlasterAS 2590玻璃钢加固石膏板AS 2591 The Erection and Fixing of Glass Fibre Reinforced Gypsum ProductsAS 2591竖立及固定玻璃钢或加固石膏板AS 2592 Gypsum Plaster for Building PurposesAS 2592用于建筑目的的石膏板AS 2573 Adhesivers – Mastic – For Bonding Gypsum Plaster Linings to Wood andMetal FramingMembersAS 2573 粘合剂–乳香–用于木制和金属框架的粘合得石膏板AS 2785 Suspended Ceilings – Design and InstallationAS 2785 吊顶–设计和安装AS 29082 Cellulose Cement Products – Flat SheetsAS 29082纤维素水泥产品平板AS 1111 Iso Metric Hexagon Commercial Bolts and ScrewsAS 1111 国际标准化组织公用的六角形商业螺钉和螺丝AS 1112 Iso Metric Hexagon Nuts including Thin Nuts Slotted Nuts and Castle NutsAS 1112 国际标准化组织公用的六角形螺母包括薄螺母有细长孔的螺母以及槽形螺母AS 1231 Aluminium and Aluminium Alloys – Anodised Coatings for ArchitecturalApplicationsAS 1231 铝和铝合金–应用于建筑的氧材料AS 1627 Metal Finishing – Preparation and Pretreatment of SurfacesAS 1627金属表面–表面的准备和预处理AS 1650 Hot-Dipped Galvanized Coatings on Ferrous ArticlesAS 1650铁制品的电镀层AS 1664 Rules for the Use of Aluminium in StructuresAS 1664铝结构的使用规则AS 1665 Welding of Aluminium StructuresAS 1665铝结构的焊接法AS 1734 Aluminium and Aluminium Alloys –Flat Sheet Coiled Sheet and PlateAS 1734铝和铝合金平板带状薄板和金属板12AS 2039 Methods for Testing Anodic Oxidation Coatings on Aluminium andAluminium AlloysAS 2039铝和铝合金阳极氧试验方法CEILING INSTALLATION GENERALLY天花板安装概要21 All ceiling designs shall be according to patterns as shown on drawings所有的天花板设计都应符合图纸上的型号22 Suspension system shall be rigid both horizontally and vertically and must allowfor fine adjustment each way吊顶系统按严格规定应是垂直的和水平的并且能够随时进行调整23 All members of surfaces shall be aligned for true level surfaces and straight lined所有挂件底表面应在一个平面上水平面或直线24 All joints between ceiling and wall surfaces are to be given extra care detailedand installed asshown on drawings天花板和墙表面所有的接合点都应特别设计并按图纸所示进行安装25 Using the Hilti-gun or equal is recommended for installing the ceiling hanger for"Cemboard"extruded aluminium and acoustic tiles and it will be shot directly below the concretefloor使用柄或推荐的类似物件来安装天花板挂件与铝合金连接并且铝合金挂件与混凝土板相连接26 Installation of ceiling tiles shall be installed under temperature and humidityconditions closely approximating those which will exist when the building is occupiedie they should only be installed after all wet trades are completed and all windowsand doors are in place and glazed and when the building is completely dry。
木材调运检疫规范流程英文回答:The process of wood transportation and quarantine involves several steps to ensure the safety and compliance of the wood products. Here is a detailed description of the standard procedure:1. Documentation and Inspection:Before the wood is transported, proper documentation must be prepared, including permits and certificates. This documentation ensures that the wood has been legally harvested and meets the required standards. Additionally, the wood is inspected by quarantine officers to check for any signs of pests or diseases.For example, when I worked as a quarantine officer, I would examine the documents provided by the wood supplier to verify their authenticity. I would also visually inspectthe wood for any visible signs of pests, such as holes or tunnels.2. Treatment and Disinfection:If any pests or diseases are detected during the inspection, appropriate treatment and disinfection measures are taken. This may involve fumigation, heat treatment, or chemical treatments to eliminate or control the pests.For instance, if I found evidence of wood-boring beetles during the inspection, I would recommend the wood to be treated with heat to kill the larvae and prevent further infestation.3. Packaging and Labeling:After the wood has been inspected and treated, it is properly packaged and labeled for transportation. The packaging should be secure and robust to prevent any damage during transit. The labels should clearly indicate the origin, treatment methods, and any other relevantinformation.As an example, when I supervised the packaging process, I would ensure that the wood was tightly wrapped in plastic to protect it from moisture and external contaminants. I would also check that the labels were affixed properly and contained all the necessary information.4. Transportation and Monitoring:During transportation, the wood is closely monitored to ensure compliance with the quarantine regulations. This may involve periodic inspections, temperature monitoring, and tracking the movement of the wood.For instance, if the wood is being transported by truck, I would conduct random inspections at checkpoints along the route to ensure that the wood is still in compliance with the quarantine requirements. I would also monitor the temperature inside the truck to prevent any moisture-related issues.5. Final Inspection and Release:Once the wood reaches its destination, a finalinspection is conducted to ensure that it has remainedpest-free during transit. If the wood passes the inspection, it is released for further processing or distribution. However, if any issues are detected, appropriate measures are taken to address the problem.For example, when I performed the final inspection, I would carefully examine the wood for any signs of pests or damage. If everything looked satisfactory, I would issue a release certificate, allowing the wood to be used or sold.中文回答:木材调运和检疫的流程涉及多个步骤,以确保木材产品的安全和合规性。
无人水面艇检验指南英文Unmanned Surface Vessel Inspection GuideIntroduction:Unmanned Surface Vessels (USVs) are an important tool in various maritime applications, including surveillance, research, and environmental monitoring. As with any vessel, it is essential to conduct regular inspections to ensure the safe and efficient operation of USVs. This inspection guide provides a general overview of the key areas to be considered when inspecting USVs.Pre-Inspection Preparations:1. Review the USV's operation manual and familiarize yourself with its components and systems.2. Ensure that all necessary safety equipment, such as life jackets, first aid kits, and fire extinguishers, are readily available and in good working condition.3. Conduct a visual inspection of the USV to check for any obvious signsof damage or wear.Inspection Checklist:1. Hull and Structure:- Check the hull for any signs of cracks, dents, or corrosion.- Inspect the external and internal structure for any damage that may compromise the integrity of the vessel.2. Propulsion System:- Examine the propulsion system, including the propellers and motors, for any signs of damage or wear.- Test the steering mechanism to ensure it is functioning properly.3. Electrical Systems:- Inspect the USV's electrical systems, including the battery, wiring, and control panels, for any signs of wear or corrosion.- Test all electrical components to ensure they are functioning properly.4. Communication and Navigation Systems:- Check the USV's communication and navigation systems, such as radios, GPS, and radar, for proper operation.- Ensure that all necessary equipment for safe navigation, such as lights and sound signals, are functional.5. Payload and Equipment:- Inspect any additional equipment or payloads carried by the USV, such as sensors or cameras, to ensure they are properly secured and operational.6. Emergency Systems:- Test the USV's emergency systems, such as the remote control shutdown and emergency buoy deployment, to ensure they are functioning properly.Post-Inspection Actions:1. Document any findings from the inspection and report any issues tothe appropriate authorities.2. Make any necessary repairs or adjustments to ensure the USV is in safe and proper working condition.3. Conduct regular follow-up inspections to ensure continued safe operation of the USV.Conclusion:Regular inspections of unmanned surface vessels are essential to ensure their safe and efficient operation. By following this inspection guide and conducting thorough and regular checks, operators can help mitigate potential risks and ensure the reliable performance of USVs in various maritime applications.。
提示语1. 洗手间由此向前(游客服务中心内)Whereby the forward toilet (Visitor Center )2. 江苏省文物保护单位:天山汉墓时代*西汉1982年3月江苏省人民政府公布扬州市人民政府立(无英语译文)3. 阙Que (watchtower)4. 卫生间Toilet5. 廊柱已朽请勿靠近(无英语译文)6. 禁止吸烟:No smoking7. 不可回收物:NO RECYCLABLE8. 可回收物:RECYCLABLE9. 灭火器箱:FIRE EXTINGUISHER BOX火警119:FIRE ALARM 11910. 王后寝宫:QUEEN’S IMPERIAL SLEEPING PLALACE11. 游客服务中心:TOURIST SERVICE CENTER12. 女洗手间female toilet13. 桧柏Chinese junipers广玉兰broad yulan magnolia香樟Incense camphor tree海桐pittosporum tobira琼花the fine jade is spent桂花fragrans14. 扬州市旅游区门票价目表(汉广陵王墓博物馆)ScenicTicket Price Lst of Yangzhou(Han Dynasty Tomb of Guangling Emperor Museum)服务项目. Service Items门票价格(人民币/元). Ticket Price(RMB/Y uan)门票. Admission Ticket收费依据. Fee-basis导游费Open hours开放时间Open hours价格监督电话Price Supervisory Telephone元/人次. Yuan/Person-time备注Remarks六周岁(含6周岁)以下儿童凭有效证件免费入园;身高1.4米(含1.4米)以下儿童免费入园;6周岁(不含6周岁)-18周岁(不含18周岁)未成年人,凭有效证件半价优惠入园Free for children under six years old (including 6) with valid documentation, free for children whose height is less than 1.4 m ( including 1.4m); juveniles from 6 years old (6 not included) to 18 years old ( including 18 ) with valid documentation can enjoy half-price tickets.购票须知Notice to Buying Tickets门票当日有效、一经售出概不退换. The admission ticket is good only for the date of issuewithout any replacement once being sold out.扬州市物价局检查分局监制. Supervised by the inspection sub-branch of Yangzhou MunicipalPrice Bureau15. 入口. Way in咨询服务Information Service投诉受理Tourist Complaints Handling行李寄存Left Luggage提供轮椅Wheel chair急救医药用品First aid medical supplies报刊阅览Newspaper &Mags游览宣传册Tour brochure无线上网WI-FI Wireless Internet WI-FI当心台阶W ATCH YOUR STEP导游服务处GUIDE SREVICE票务服务TICKET OFFICE入口ENTRANCE出口EXIT医疗点FIRST AD投诉办公室COMPLAINT OFFICE游客服务中心TOURIST SERVICE CENTER卫生间Toilet灭火器fire extinguisher小心地滑caution wet floor广陵潮涌tide in guangling扬州最中国find china in yangzhou咨询、投诉受理inquiries & complaints accepting禁止吸烟no smoking不乱扔垃圾不随地吐痰可回收墓内介绍:1. 饮食用具:YINSHIYONGJU“食官第一内户”出土了成组的饮食用具,有陶壶、陶瓿、铜鼎、漆案、漆盘、漆碗、成套的漆耳杯等……这些漆制餐具的出土,反映了漆器在上流社会的广泛使用和饮食文化的丰富多彩。
施工现场材料、设备进场报验管理流程1.施工现场材料、设备进场前需进行报验。
Materials and equipment need to be inspected before entering the construction site.2.报验包括材料、设备的规格、数量和质量等信息。
Inspection includes information such as specifications, quantity, and quality of materials and equipment.3.填写报验单,并由相关负责人签字确认。
Fill out the inspection form and have it signed by the responsible person.4.报验单需附上材料、设备的相关证书和检测报告。
The inspection form should be accompanied by relevant certificates and test reports for materials and equipment.5.相关人员对材料、设备进行外观检查和质量把关。
Relevant personnel conduct visual inspections and quality control of materials and equipment.6.确认材料、设备符合施工标准和要求后方可进场。
Materials and equipment can only enter the site after confirming that they meet construction standards and requirements.7.每批材料、设备进场均需进行报验,不得擅自放行。
Each batch of materials and equipment entering the site must be inspected and cannot be released without authorization.8.报验人员需将报验情况及时记录在报验台账中。
澳大利亚进口再造木制品、竹木草制品的检疫要求再造木制品(reconsituted wood products)Commodity: Reconstituted wood productsScientific name:Country: All countriesEnd use: All uses other than as animal foods, fertilisers or for growing purposes产品:再造木制品来源国:所有国家用途:除饲料、化肥及种植以外的所有用途以下为具体检疫要求Non-Commercial1. The conditions under the Commercial section apply非商业性的货物,检疫措施与商业性再造木制品相同Commercial1. An Import Permit is not required.2. A Quarantine Entry is not required.3. Reconstituted wood products are not of interest to quarantine as they are deemed to pose minimal quarantine risk and can be released without inspection. Reconstituted wood products are those that as a result of the manufacturing process no longer contain solid wood and include particleboard, chipboard, hardboard (masonite), oriented strand board, medium and high densityfibreboard. This case does not apply to plywood and veneer products. (Refer to the ICON commodity ‘Plywood and Veneer’.)商业性的货物1. 不需要进境许可。
Olli Silvén, Matti Niskanen, and Hannu KauppinenThe appearance of sawn timber has huge natural variations that a human inspector easily compensates for in his brain when determining the types of defects and the grade of each board. However, for automatic wood inspection systems these variations are a major source for complication. For instance, normal wood grain and knots should be reliably discriminated in all circumstances, but simple thresholding based detection methods used in state-of-the-art inspection systems frequently fail even in this apparently straightforward task. Furthermore, the appearances of defects vary, so one board’s sound knot is another board’s dark one, making it difficult to use textbook methodologies for visual inspection. These generally aim at systems that are trained in a supervised manner with samples of defects and good material, but selecting and labeling the samples is an error prone process that limits the accuracy that can be achieved.We have developed a non-supervised clustering based approach for detecting and recognizing defects in lumber boards. The solution is simple to train, and supports detecting knots and other defects by using multidimensional feature vectors containing texture and color cues for small (e.g., 32-by-32 pixel) non-overlapping regions in the image. A key idea is to employ a Self-Organizing Map (SOM) for discriminating between sound wood and defects. An identical scheme is employed in classifying the defects. Human involvement needed for training is minimal.The approach has been tested with color images of lumber boards. Based on visual judgment, the false detection and error escape rates are very low. The defect classification accuracy is harder to assess, because of the non-supervised principle of operation, however, it is not affected by human errors in training and is thus probably better. The approach also provides a self-intuitive visual user interface.Department of Electrical Engineering, P.O.B. 4500, FIN-90014 University of Oulu E-mail: {Olli.Silven, Matti.Niskanen, Hannu.Kauppinen}@ee.oulu.fiThis paper is available at http://www.ee.oulu.fi/~olli/Projects/Lumber.Grading.htmlThe visual grading of wood surfaces is a problem that combines very formal grading rules, large natural variations in the target material, and highly subjective appearance criteria. These factors make it a very challenging task, which is not ideal for humans that have difficulty in following the strict and detailed grading standards. These consider every single defect type, size, and position, while human classification seems to be based on an overall impression, augmented with some specific rules, such as the presence of knots at the edges of the board. On the other hand, a human inspector seems to painlessly compensate for the variations in the material in his brain, while this part of the task seems to be particularly difficult for machine vision systems used for automatic grading.Human visual inspection rarely achieves better than 70% performance in grading lumber. Huber et al (1985) found that in grading red oak lumber the human inspectors achieved 68% of the perfect result when recognizing, locating, and identifying defects. Grönlund (1995) found that when grading boards into four grades, only 60% of the boards were assigned the same grade by two different expert inspectors.In automatic visual inspection it is not straightforward to determine the actual accuracy achieved, as the results need to be compared to the ones of humans. Lampinen et al (1994) in their comparative study found that knots in pine boards could be classified at 80% accuracy. The test material consisted of about 400 high quality color images of knots classified by an expert. Kauppinen (1999) has determined that this result corresponds to 80% accuracy in grading boards when Nordic Timber (1994) rules are enforced. We believe this is the best that can be achieved with state-of-the-art automatic inspection systems.Unfortunately, the performance of inspection systems suffers from the variations in the material, e.g., normal wood grain and knots may be difficult to discriminate in all circumstances. A big part of the blame falls onto the used popular thresholding based detection methods. In addition, the appearances of defects vary, so one board’s sound knot is another board’s dark one, requiring contextual information for proper classification. Figure 1 shows six knots classified as sound and another six classified as dry by a human expert. The range of variation within classes is clearly substantial.a)b)Examples of knots: a) Sound knots, b) dry knots. Within classes the variation is large, e.g., the color of some knots can be lighter than background of others.The textbook methods proposed for visual inspection generally aim at systems that are trained by showing samples of defects and sound material as presented by Figure 2. However, we believe this supervised training approach is fundamentally flawed in wood inspection. Selecting and labeling the samples is an error prone process that limits the accuracy that can be achieved, and this principle is not easy to make adaptive to cope with the variations of material. The result is more likely to be a frustrating everlasting training task rather than an ideal accurate automatic visual inspection device. Not surprisingly, current state-of-the-art systems seem to employ rule based classification that is a rather rigid approach.The block diagram of visual surface inspection using superviced classification.In this paper we present a unified solution to the shortcomings in detection and classification of defects of lumber boards, describe experiments, giving some quantitative data on performance, and discuss the consequences for inspection system design and principles of grading. The proposed approach relies on non-supervised training and does not require labeling individual samples, but uses a clustering method to discover whether the samples fall into a finite set of categories based on their similarity relations. Thus, it is not affect by human errors in training. The approach also provides a self-intuitive visual user interface that is very easy to understand even for novices in the field.In our approach, wood inspection is divided into two phases. In the first one we detect possible defected regions. In the second one these are examined as individual defects. Classification in both phases are based on non-supervised clustering, implemented using self-organizing maps (SOMs). For the reader of this paper it suffices to understand that the SOMs project a multidimensional feature space into a two-dimensional presentation.Most of the features used in our experiments are centiles that are color histogram features introduced by Silvén and Kauppinen (1994). These work very well with highly varying lumber images. In addition, we have experimented with second order texture features, such as Local Binary Patterns (LBP) (Ojala et al. 1996) and statistical features calculated from co-occurrence matrixes (Haralick & Shapiro 1992). Centiles are only marginally more expensive to calculate than the trivial adaptive gray level thresholding methods used in most automatic lumber grading systems.The test material used in our experiments has been prepared by VTT Building Technology. It consists of a large number of pine boards with ground truth classifications for each defect. The results reported here are based on a set of 20 boards totaling around 400 MB of data. The imaging resolution has been 0.5 mm and a color line-scan camera has been used for image acquisition.The defect detection algorithm used is shown in the diagram of Figure 3. The wood image captured by the imaging system is divided into non-overlapping rectangular regions of a fixed size (e.g. 32*32 pixels). A set of feature values, in our case centiles of color histograms, is calculated for each region, resulting in a feature vector. Each feature vector is then mapped into a self-organizing map that we call .The defect detection algorithm. In the SOM at the bottom of the figure, the distance between adjacent nodes represents their relative distance in a high dimensional feature space. The dashed line is a manually drawn border between sound and defected wood.In the SOM, for each feature vector, a node that is closest in the feature space is chosen and the corresponding sample is added to the respective cluster. The result is a 2D map in which regions with similar feature vectors are located close to each other. The dashed line in the SOM is the boundary between sound wood and suspected defects. This boundary is set during the basic training stage of the algorithm and is the only human interaction required. In practice, this defect detection approach can be interpreted as thresholding in a multidimensional feature space.An example of a SOM used for defect detection from pine boards is shown in Figure 4. For each node one randomly selected sample is shown. In our experimental system, clicking any node with a mouse results in showing all the samples clustered to the node. The gray nodes have no samples. The similarities between adjacent nodes are apparent. Dry knots and sound wood are at opposite corners of the SOM while the splits and other defects are in-between. Two different approximations of boundary between sound wood and suspected defect regions are shown in the figure.Detection SOM: The dashed lines are manually drawn approximations of boundary between sound wood and defects.In our experiments, setting a fixed border between sound and defected wood has given good results, despite material variations, and only minor improvement has been achieved using adaptive, that is, board-specific boundaries. The boundary can be determined quite easily by just taking a glance at the SOM. Figure 5 shows what happens with two different selections, optimistic and pessimistic, for the boundary.With the optimistic selection the defect regions detected are quite small, while the pessimistic selection includes even the transitions from sound wood to defects in thedetections. The number of false alarms is minimal in both cases.a) Pessimistic and b) optimistic selection of the boundary. The boundariesused are sketched in Figure 4.a)b)In practice, we favor pessimistic boundaries as they result in regions that include information of the background of the defect. This improves the classification results as the knots with same color but with different kinds of backgrounds are clustered in to their own nodes in the detection SOM, see Figure 1 for examples of such knots.The defect classification method is schematically depicted in Figure 6. The defect detection stage results in set of elementary suspicious regions. These are now subjected to connected components analysis to obtain regions for the recognition phase. For each of these connected areas a new set of features is calculated. The previously calculated centile features are used in our experiment here, although different centiles may be more powerful in discriminating between defects than for separating sound wood and defects.Recognition algorithm. Defected areas form regions that are clustered to SOM as individual defects.Suspected defect regions fromHistograms Additional features(e.g. principle axes)CentilesFeature vector To RAt this stage only a very small fraction of the original data remains enabling us to use more sophisticated and computationally complicated methods for analysis. The category of a knot depends on its aspect ratio, and its size is also important in grading rules. The method we use for calculating the orientation and lengths of the major and minor axes of an ellipse is based on the first and second order moments (Haralick & Shapiro 1992). Although all the blobs are not ellipsoidal, an ellipse can approximate quite well the convex shape of most defects in wood. Furthermore, the location of knots is very important criteria in grading, but we have not yet used this information in our experiments.Again, the feature vectors are fed into a SOM that clusters similar defects close to each other. Figure 7 shows the result when the detections from several boards have been used to produce the . Only one randomly chosen suspected detection is shown in each node.Classification SOM. Defects with similar appearance are close to each other.It is essential to realize that so far we have not labeled any defects nor collected any special training material. Neither have we used any special rule based knowledge about defects.Evaluating the classification accuracy of this approach is not straightforward. When we take a look at the clusters at any individual node in the SOM we see that the defects are very similar and clearly belong to same category as Figure 8 demonstrates.However, when we compare them with the corresponding independent manual labelings, prepared during the material acquisition stage, there is about 20% disagreement. This is in agreement with to the results compiled by Lampinen et al (1994). In other words, if manually labeled samples are used in training an inspection system, it is difficult to achieve better than 80% accuracy. We believe that minimal human involvement in training is a key to high accuracy.a)b)c)Defects mapped to three different SOM nodes. A) dry knots, b) sound knots and c) splits.Although the methodology we are proposing for wood inspection is based on non-supervised clustering, it can not be applied in a blind manner, but a sound strategy is needed in preparing the detection and recognition SOMs. In the next sections we describe the approaches found powerful in building the SOMs.A fundamental problem in visual inspection is the selection of efficient features for defect detection and recognition purposes. In practice, these features need to be searched using test material that has been prepared in a supervised manner. In our case this very time consuming process was performed for a small amount of material using a Knn-classifier, and achieved an 80% recognition accuracy.Only a fraction of the surface area of a wooden board contains defects. If we simply divide the image of the board into, say, 32*32 pixel regions, calculate the features and build the SOM, the result has the defects concentrated into a few nodes at one of the corners while the remaining nodes contain sound wood regions. This prevents the straightforward application of SOM methodology, because determining the boundary between sound wood and defects is difficult. In practice, it would be beneficial to have wide transition regions in the SOM between different categories of wood as the discrimination result would then be more robust with respect to location of boundary.Because of this need, the defect detection SOM is built using selected test material that contains about the same amount of sound wood and defects. This training material originates from several boards to capture the whole range of variation of lumber. The selection process does not require any labeling of the samples, nor any particular expertice in choosing the material.In order to obtain an ideal recognition SOM we should have around the same number of every defect category in the sample set, because the under-represented categories may not get a node of their own, while the larger classes conquer most of the map. To cope with this phenomenon we must give some of the samples more weight by using them multiple times when building the SOM. In practice, teaching the SOM is an iterative process in which adding weight to some samples causes the whole map to change and brings up some other nodes whose weights should then be adjusted. Clearly, an efficient tool is needed for this purpose.We use a software package, called G-SOM (1999), to create self-organizing maps. It provides visual support for finding which samples are clustered in each node, so one can easily modify the training data and see the discriminative power of the approach. This tool enables to label each node and the respective samples manually in a very efficient manner, so it can be used for preparing training material for supervised classification solutions, if desired.The weak points in automatic visual inspection of lumber are human involvement in system training and overly simplistic defect detection methods. The subjective judgment of human appears to be in conflict with precise, formal grading rules and defect criteria. The decision of a human on the category of any individual defect is inherently error-prone and should thus be avoided, however, a human can easily label a group of defects into a single category. This is exactly the kind of approach what the presented solution supports. The defect detection solution works in the same manner, avoiding too early decisions on the type of the regions.The support of TEKES for this line of research is gratefully acknowledged. We also thank Arto Usenius and Olof Sommardahl at VTT Building Technology, and Hannu Rautio and Toni Piirainen at the University of Oulu for their invaluable contributions.Grönlund U (1995) Quality improvements in forest products industry. Dissertation, Luleå University of Technology, Sweden.G-SOM tool (1999), University of Oulu, Department of Electrical Engineering. URL: http://www.ee.oulu.fi/research/imag/gsom/Haralick R & Shapiro L (1992) Computer and robot vision, vol.1, Addison-Wesley. Huber HA, McMillin CW & McKinney JP (1985) Lumber defect detection abilities of furniture rough mill employees. Forest Products Journal 35: 79-82.Kauppinen H (1999) Development of a color machine vision method for wood surface inspection, Dissertation, Acta Universitatis Ouluensis C141 University of Oulu, Department of Electrical Engineering, Oulu, Finland, 138p.Lampinen J, Smolander S, Silvén O & Kauppinen H (1994) Wood defect recognition: a comparative study. Workshop on Machine Vision for advanced production, Oulu, Finland.Nordic timber, grading rules (1994) Pohjoismainen sahatavara, lajitteluohjeet (in Finnish). Suomen Sahateollisuusmiesten Yhdistys, 64p.Ojala T, Pietikäinen M & Harwood D (1996) A comparative study of texture measures with classification based on feature distributions. Pattern Recognition 29:51-59. Silven O & Kauppinen H (1994) Color vision based methodology for grading lumber.Proc. 12th International Conference on Pattern Recognition, Vol. 1, October 9-13, Jerusalem, Israel, pp. 787-790.。