Pattern Recognition 40(2007)309–
320
https://www.doczj.com/doc/c98445987.html,/locate/patcog
Af?ne invariant comparison of point-sets using convex
hulls and hausdorff distances
C.Gope,N.Kehtarnavaz ?
Department of Electrical Engineering,EC 33,University of Texas at Dallas,Richardson,TX 75083-0688,USA
Received 31October 2005;received in revised form 3March 2006;accepted 24April 2006
Abstract
Many object recognition or identi?cation applications involve comparing features associated with point-sets.This paper presents an af?ne invariant point-set matching technique which measures the similarity between two point-sets by embedding them into an af?ne invariant feature space.The developed technique assumes no a priori knowledge of reference points,as is the case in many identi?cation problems.Reference points of a point-set are obtained based on its convex hull.An enhanced version of the Modi?ed Hausdorff Distance is also introduced and used in the feature space for comparing two point-sets.It should be noted that the technique does not attempt to obtain correspondences between the point-sets.The introduced technique is applied to two real databases and its performance is found favorable as compared to three other af?ne invariant matching techniques.
?2006Pattern Recognition Society.Published by Elsevier Ltd.All rights reserved.
Keywords:Af?ne invariant;Convex hull;Hausdorff distance;Point-pattern comparison;Shape matching
1.Introduction
Object matching is required in many recognition or iden-ti?cation applications.A set of features is usually extracted from a query image and matched against a database of im-ages with similar features in order to identify the best pos-sible match.Among the most commonly used features are points,lines,and curves.Of course,the suitability of the cho-sen features depends upon the application.For example,in [1,2],we used the curve features for the photo-identi?cation of marine mammals.There are other applications,where point-sets provide the most reliable or the only available features.This work addresses the problem of measuring the similarity of point-sets under possible af?ne transformations,a widely encountered problem in diverse ?elds such as com-puter vision,pattern recognition,computational geometry,molecular biology,etc.
Although there are applications for which the knowledge of a few distinguishing reference points is available,the problem studied here assumes no such a priori knowledge.It
?Corresponding author.Tel.:+19728836838;fax:+19728832710.
E-mail address:kehtar@https://www.doczj.com/doc/c98445987.html, (N.Kehtarnavaz).
0031-3203/$30.00?2006Pattern Recognition Society.Published by Elsevier Ltd.All rights reserved.doi:10.1016/j.patcog.2006.04.026
is shown that for a point-set with irregular (non-symmetric)point distributions,af?ne invariant reference points can be generated from the point-set itself.The introduced af?ne invariant matching technique utilizes the convex hull of the point-set to extract af?ne invariant features.Variations of the Hausdorff distance are then used for comparing two point-sets embedded in the af?ne-invariant feature space.A major attribute of this matching technique is that it does not require any user-de?ned parameters.The matching results are shown for two real databases and the performance is compared to three af?ne matching techniques,namely Af?ne Moment Invariants,Discrete Af?ne Moments,and Invariant Feature Vector using convex hulls.Noise and occlusion effects are also studied and compared.
The paper is organized as follows.Section 2provides an overview of the previous work as related to point-set com-parison and matching.Section 3brie?y describes the prob-lem of af?ne invariant matching.Then,Section 4describes the developed af?ne invariant point-set matching technique using convex hulls and Hausdorff distances.Experimental results are presented in Section 5.Finally,the paper is con-cluded in Section 6.
310 C.Gope,N.Kehtarnavaz/Pattern Recognition40(2007)309–320
2.Previous work on point-set matching
A number of techniques addressing the problem of point-set or point-pattern matching have appeared in the literature. Clustering approach is used in[3]to simultaneously deter-mine the transformation(restricted to translation,rotation and scaling)and point matching between two images.Trans-formation parameters for all possible point-pairings are cal-culated and then clustered in a parameter space,with the strongest cluster representing the most likely transformation. Such methods are not computationally ef?cient which lim-its their usage.It is worth pointing out that the technique described in this paper generates af?ne invariant reference points which could be effectively used to improve the ef-?ciency of such clustering approaches.In[4],a faster2D clustering approach is introduced where the matching is in-variant to translation,rotation,and scale changes.In[5],a statistical framework for iterative alignment and correspon-dence of point-sets is discussed with respect to two differ-ent models,the Procrustes model and the point-distribution model.In[6],a least-squares estimation approach invariant to translation,rotation,and scale changes is discussed for two point-sets having the same cardinality.Inter-point dis-tances are used in[7,8]for point-set comparison employing graphs and search trees.In[9],inter-point distances are cal-culated for all the points in a point-set and then some heuris-tics are used to obtain local matches between two point-sets. This process is iterated until an acceptable global match is obtained.In[10],a geometric alignment strategy is used while the concept of geometric hashing is utilized in[11]. Hausdorff distances for image matching invariant to trans-lation,rotation,and scale changes are presented in[12].In [13],the Hausdorff distance method is extended to the more general af?ne transformation for locating objects in a scene. In this method,a grid is imposed on the space of possible af?ne transformations and the complexity of the algorithm is dependent upon the restrictions applied to the allowed trans-formations.In[14],a hierarchical top-down approach is used to estimate the best aligning af?ne transformation among all possible transformations.The Iterative Closest Point algo-rithm is introduced in[15],where an iterative least-squares technique is used to obtain3D motion from the point correspondences between two point-sets.In[16],matching is performed by?nding the correspondence between af?ne transformed point-sets where all the possible sets of four points in a set are taken into consideration.Fourier descrip-tors are used in[17]for af?ne invariant recognition of3D objects.In[18],a point matching strategy is discussed using af?ne invariant representation of points based on a triplet of basis points.All possible triplets are considered and the af?ne invariants of the remaining points are computed and stored in a hash-table,which is then used in the matching stage.In[19],an explicit noise model and an optimal voting approach are used to achieve a robust af?ne invariant point matching.In[20],a global af?ne transform correlation is used to align af?ne transformed gray-level images.
Convex hull edges of a point-set are utilized in[21]which
serve as the features for similarity invariant(translation,ro-
tation and scale changes)point matching.Viewpoint invari-
ant Fourier descriptors in combination with convex hulls are
presented in[38]for similarity invariant shape matching.
In[23],af?ne invariant representations of point-sets are ob-
tained by using distance ratios de?ned by quadruples of fea-
ture points.Then,the convex hull of a point-set is utilized to
select some reference points.Vertices of the convex hull of
a point-set are used for af?ne invariant point-set matching
in[22].Af?ne invariants are constructed using four consec-
utive vertices of a convex hull at a time.These invariants
are then used to estimate a global aligning transformation
between two point-sets.The performance of this technique
has been compared to that of the technique introduced in
this paper,since it also uses the convex hull of a point-set to
derive af?ne invariants for matching.This technique is re-
ferred to as the Invariant Feature Vector(IFV)in the results
section.
Moments are widely used in point-based object recog-
nition and alignment.In[24],cross-weighted moment in-
variants are used for alignment and recognition.Af?ne mo-
ment invariants of point-sets are used in[25,26]while Dis-
crete Af?ne Moments are used in[27,28]for af?ne invariant
matching of discrete point-sets.It should be noted that Af?ne
Moment Invariants are derived using the algebraic theory
of moment invariants whereas Discrete Af?ne Moments are
derived via the method of normalization.We have also com-
pared the matching performances of these two techniques to
that of the technique introduced in this paper.Very recently,
Af?ne Moment Descriptors are discussed in[29]as an ex-
tension to the technique covered in[30],in order to estimate
the aligning transformation between two point-sets,where
it is shown that the resulting descriptors can be converted to
Af?ne Moment Invariants.
3.Af?ne transformation and invariants
Af?ne transformation is an important subgroup of the
general class of projective transformation.It includes trans-
lation,rotation,scaling,and skewing.In many vision and
object recognition applications,it is necessary to consider
invariance to af?ne transformations because images are
captured not only from different distances but also with
different degrees of out-of-plane rotations.Also,it is suf-
?cient to consider af?ne invariance(instead of the more
general projective invariance)if the object size is small
as compared with the distance between the object and the
camera.
Formally,the problem of af?ne invariant point-set or
point-pattern matching can be stated as follows:given
two points-sets U={u i}m i=1and V={v i}n i=1,possibly re-lated by an af?ne transformation,compute a similarity(or
equivalently,dissimilarity)measure between the point-sets.
Mathematically,for2D points,an af?ne transformation can
C.Gope,N.Kehtarnavaz /Pattern Recognition 40(2007)309–320311
be stated as
v x
v y
=T 2x 2 u x u y +b 2x 1,(1)
where v x v y
represents the transformed coordinates of an original point u x
u y ,matrix T represents rotation,scaling,and skewing transfor-mations,and vector b represents translation.If T is a full-rank matrix,then the af?ne transform maps 2D points into 2D points,with the area of the transformed object being scaled by a factor |T |,where |·|denotes matrix determinant.3.1.Af?ne invariants:barycentric coordinates
A set of points can be used to create an af?ne frame.Consider a triangle R 1R 2R 3,and a point P lying in the plane of the triangle.We can write P as P = R 1+ R 2+ R 3,(2)
such that + + =1.
(3)The coordinates ( , , )can be uniquely obtained and are called the homogeneous barycentric coordinates [31]of the point P with respect to R 1,R 2,R 3.From Eqs.(2)and (3),the following matrix equation can be written R 1x R 2x R 3x R 1y R 2y R 3y 1
1
1
= P x P y 1
.(4)
Using the Cramer’s rule for a system of linear equations,it can easily be shown that if the area of the triangle R 1R 2R 3is non-zero,i.e.R 1,R 2,R 3are not collinear,then we have
=
P x R 2x R 3x P y R 2y R 3y 111
R 1x R 2x R 3x R 1y R 2y R 3y 111 ; = R 1x P x R 3x R 1y P y R 3y 111
R 1x R 2x R 3x R 1y R 2y R 3y 111 ;
= R 1x R 2x P x R 1y R 2y P y 111
R 1x R 2x R 3x R 1y R 2y R 3y 111
.
(5)Note that when Eq.(3)is true,the combination R 1+ R 2+ R 3is called an af?ne combination and any af?ne transformation preserves this combination.Thus,given three reference points in a plane,any other point in the plane can be described by its barycentric coordinates ( , )and this representation is invariant to af?ne transformations ( is not needed as it can be obtained from and ).
4.Matching using convex hull and hausdorff distances In this section,we introduce our af?ne invariant point-set matching technique which does not assume any a priori knowledge of reference points.This approach uti-lizes the convex hull of a point-set,and its properties,to generate an af?ne invariant representation of the points.First,let us start with a brief introduction to convex hulls.
4.1.Convex hull
The convex hull of a point-set is the smallest convex space
that contains the points.For a ?nite 2D point-set,the convex hull is the smallest convex polygon containing all the points.Fast algorithms exist for computing convex hulls.Here,we have used the Quick Hull algorithm [32].The worst-case complexity of this algorithm for a point-set containing n points is O(n log n).
Convex hulls have some useful properties that make them suitable for many recognition and representation tasks.Apart from their computational ef?ciency ,convex hulls are par-ticularly suitable for af?ne matching as they are af?ne in-variant [33].In other words,if a point-set undergoes an af?ne transformation,the convex hull of the point-set un-dergoes the same af?ne transformation.Also,convex hulls have local controllability ,i.e.they are only locally altered by point insertions/deletions/perturbations.This is a useful property as far as noise tolerance and partial occlusions are concerned.
4.2.Convex hull based af?ne invariants
Consider a point-set ={P 1,P 2,...,P k ,...,P n }con-sisting of n two-dimensional points.It is desired to obtain an af?ne invariant representation of .As previously pointed out,the barycentric coordinates ( , )of a point provide an af?ne invariant representation.In order to compute the barycentric coordinates,three reference points need to be available.In what follows,it is shown how such reference points can be obtained from the point-set itself and the con-vex hull of the point-set.
Let C H denote the convex hull (polygon)for the set and let H be the set of L points (L n)constituting the convex hull C H ,and = ? CH be the set of points in inside the convex hull.Let (x i ,y i ),i =1,2,...,L ,be the ordered vertices forming the polygon C H .Using the Green’s
312 C.Gope,N.Kehtarnavaz /Pattern Recognition 40(2007)309–320
theorem,it can be shown (for proof,refer to Appendix A)that the area A of C H is given by A =12
L
i =1(x i y i +1?x i +1y i ).
(6)
Also,the centroid of C H denoted by (c x ,c y )can be ex-pressed (for proof,refer to Appendix B)as c x =16A L
i =1(x i +x i +1)(x i y i +1?x i +1y i ),
c y =
16A
L i =1(y i +y i +1)(x i y i +1?x i +1y i ).
(7)
In Appendix C,it is proved that the centroid (c x ,c y )is af?ne invariant,i.e.the centroid of the af?ne transformed convex hull is the af?ne transformed centroid of the original convex hull.It can easily be shown that the mean (m x ,m y )of the set is also af?ne invariant,i.e.the mean of the af?ne transformed points of is the af?ne transformed mean of the original points of .Thus,the centroid of the convex hull and the mean of all the points in the point-set (indicated by R 1and R 2in Fig.1,respectively)can serve as two reference points for the points in the set ,with the assumption that they are not coincident.This assumption is generally true for irregular shapes and point patterns encountered in many real-world problems,such as the marine mammal recognition problem presented later in the results section.Thus,it
should
Fig.1.Selection of three reference points (R 1,R 2and R 3)for a feature point P k .
be noted that the introduced matching scheme should not
be used for datasets where R 1and R 2are coincident.In practice,however,such datasets rarely occur.
Having obtained R 1and R 2,Fig.1shows how the third reference point R 3is located.The line P k R 1is extended to intersect the convex hull at Q .Similarly,P k R 2is extended to intersect the convex hull at S .The point R 3is the mid-point of the line segment QS .By using the well-known properties:(a)the intersection point of two straight lines is preserved
under af?ne transformations,and
(b)the mid-point of a line segment is af?ne invariant,
it is easy to see that R 3is also af?ne invariant for the given point P k .Thus,this way,three reference points R 1,R 2,R 3for the point P k are identi?ed based on which the af?ne invariants ( , )are derived as described in Section 3.1.It should be noted that unlike R 1and R 2,the third reference point R 3depends on the feature point P k for which the af?ne invariants need to be computed (note that P k belongs to the set and not ).At this point,it should be emphasized
that in order for P k and P k (transformed P k
)to be mapped to the same point in the feature-space,it is required that
R 3and R 3(transformed R 3
)be related by the same af?ne transformation that relates R 1and R 1(transformed R 1
),and R 2and R 2(transformed R 2).This requirement is met be-cause it has been shown that (in non-degenerate cases)R 3is
af?ne-invariant for the given point P k .Hence,although R 3is different for each P k ,for a given P k ,it is af?ne invariant,
which means P k and P k
are mapped to the same point in the feature-space.
As an example,consider the point-sets shown in Fig.2.On the left is the original point-set consisting of 20points,generated randomly (uniform distribution)within a 40×40grid.Seven of its points (indicated by ‘o’)constitute the con-vex hull and the invariants for the remaining thirteen points (indicated by ‘*’)are computed using the convex hull and the reference points R 1,R 2,R 3.The point R 3is not shown in this ?gure as its location depends on the point for which the invariants are computed.On the right,an af?ne transformed version of the point-set is shown,with the transformation
T = 0.8660.1930.5001.266 and b = 55 (see Eq.(1)).
Zero mean Gaussian noise with variance 2was also added to the x and y coordinates of the transformed points.The average errors introduced in the values of and were 0.06and 0.04,respectively.Also,the maximum errors introduced in the values of and were 0.11and 0.09,respectively.Table 1shows ?ve of the invariant values before and after the transformation.As can be seen from this table,and the average and the worst-case errors reported above,the values mostly retain invariance after the transformation and noise addition.
Here,it is worth mentioning that in [22],the use of af?ne invariants based on the vertices of a convex hull is discussed,
C.Gope,N.Kehtarnavaz /Pattern Recognition 40(2007)309–320
313
Fig.2.Convex hulls (shown as bounding polygons)and reference points R 1and R 2for a point-set before (left)and after (right)an af?ne transformation.
Table 1
Af?ne invariants before and after af?ne transformation and noise addition
(original)
(af?ne transformed plus noise)
(original)
(af?ne transformed plus noise)
0.910.900.840.840.710.760.940.930.730.700.740.750.980.920.690.650.990.980.95
0.95
where a total of 10invariants for a quadruplet (4consecu-tive vertices)are utilized.However,from the theory of al-gebraic invariants [34],it should be noted that a set of n points can only have 2n ?6independent invariants as far as af?ne transformations are concerned.Hence,for n =4,there are only 2independent invariants.As a result,the other invariants that are used in [22]do not provide any extra information.
4.3.Hausdorff distance
In this section,we discuss the problem of comparing point-sets using various Hausdorff distances.Previously,Hausdorff distances have been used for point-set compari-son for the group of transformations limited to translations or rigid motions (translation and rotation).In [12],Hutten-locher et al.have proposed fast algorithms for the com-putation of Hausdorff distances,provided that point-sets lie on an integer grid.The complexity of matching grows with more general transformation groups such as af?ne transformation [13].
Given two point-sets U ={u 1,u 2,...,u m }and V ={v 1,v 2,...,v n },the Hausdorff distance is de?ned as H (U,V )=max (h(U,V ),h(V ,U)),
(8)
where
h(U,V )=max u ∈U min v ∈V
u ?v
(9)
and . is a norm de?ned on the point-set,such as the L 2norm.To cope with occlusions and outliers,partial Haus-dorff distances are introduced in [12],where the K th ranked distance is considered,instead of the maximum.In [35],the modi?ed Hausdorff distance (MHD)is utilized,in which h(U,V )is de?ned as h(U,V )=1m
u ∈U
d(u,V ),
(10)
where
d(u,V )=min v ∈V
u ?v
(11)
and H (U,V )is the same as in (8).In essence,MHD uses the average of the minimum distances,rather than the max-imum (or K th).In [35],it has also been shown that MHD has a higher discriminatory power as compared to the other Hausdorff distances.
It should be noted that the above Hausdorff distances do not incorporate any global correspondence information be-tween two point-sets.In other words,these distances do not take into consideration whether or not multiple matches are
314 C.Gope,N.Kehtarnavaz /Pattern Recognition 40(2007)309–
320
Fig.3.Modi?ed Hausdorff Distances (MHD)and Enhanced Hausdorff distances (EHD)for two point-sets indicated by ‘*’and ‘o’.
found for the same point.This can serve as a major advan-tage as it is not required to solve the often dif?cult problem of correspondence.However,at the same time,it can be a drawback as no attempt is made to utilize the correspondence information.In this paper,we have addressed this issue by incorporating some pairing information,without demanding to explicitly solve the correspondence problem.Here,we refer to our version of the Hausdorff distance as enhanced Hausdorff distance (EHD).4.3.1.Enhanced Hausdorff distance EHD is de?ned as
H EH (U,V )=max (h EH (U,V ),h EH (V ,U)),(12)
where h EH (U,V )=
1m ?
u ∈U
d(u,V ).
(13)
The term re?ects the correspondence information and d(u,V )is the same as the one de?ned in (11).To determine ,the following procedure is used:Algorithm-Determine in Eq.(13):
1.Have an array of n (equal to the cardinality of set V )cells,with each cell initialized to the value ?1.
2.For each u ∈U ,?nd the nearest point in V and increment the corresponding cell value in by 1.
3.Locate the cells in whose values are greater than 0.The term in Eq.(13)is then considered to be the sum of all these values.It should be realized that step 2does not add any additional computational burden as the nearest point needs to be found anyway for computing d(u,V ).The minimum and maxi-mum values for are 0and m ?1,respectively.The
value
Fig.4.Four logo images.
0corresponds to the case when the pairing between the two point-sets is perfect (no competing points).As a result,EHD becomes the same as MHD.The value m ?1corresponds to the extreme case where all the points in U have the same nearest point in V .It is easy to see that EHD has a higher discriminatory power than MHD,owing to the explicit use of the correspondence information.
To illustrate this point with examples,consider the two point-sets shown in Fig.3.The image shown on the right is the same as the image shown on the left,the only differ-ence is the addition of one point,shown by the arrow and labeled by ‘o’.As shown in this ?gure,the MHD for both of the point-sets is 1.4and hence one cannot discriminate between them.On the other hand,the EHD for the point-set shown on the right is slightly higher,owing to the additional unmatched point.
As another example to illustrate and compare the discrim-inatory power of MHD and EHD,consider the four logo images shown in Fig.4.Tables 2and 3list the MHD and EHD,respectively.It can be observed that the logo images 1and 3are most similar,as indicated by their MHD and EHD having the value 1.However,the ratio of the maximum to the minimum distance is 13for the EHD while it is only 11for the MHD,owing to the higher discriminatory power of EHD.Bear in mind that the higher discriminatory power of EHD may not always be desirable for the application under consideration.For example,in some applications,it
C.Gope,N.Kehtarnavaz/Pattern Recognition40(2007)309–320315
Table2
MHD for the logo images
Modi?ed Hausdorff Distance(MHD)
Image1Image2Image3Image4 Image103111
Image230314
Image31309
Image4111490
Table3
EHD for the logo images
Enhanced Hausdorff Distance(EHD)
Image1Image2Image3Image4 Image103113
Image230416
Image314012
Image41316120
may not be desired to distinguish between the two point-sets shown in Fig.3,in order to impart more tolerance towards occlusion.
4.4.Matching using convex hull based af?ne invariant features
In this section,we describe how to match two point-sets, related(possibly)by an af?ne transformation.EHD is used for matching the point-sets,with the distances computed in the feature space,as opposed to the input(point coordinates) space.It is well known[12]that under a transformation group G,the minimum Hausdorff distance M G between two point-sets is given by
M G(U,V)=min
g∈G
H(U,gV).(14)
Notice that the transformation g∈G can be applied to either U or V.Fast algorithmic solutions to(14)are dis-cussed in[12]for matching raster image data,i.e.,dis-crete grid points,when G represents a rigid transformation group.However,the complexity of the algorithm grows as invariance is sought for more general af?ne transformations [14,13].
We address this problem by computing the Hausdorff dis-tance in the feature space,which is invariant under the group of af?ne transformations.This yields a very ef?cient algo-rithm for af?ne invariant matching while still utilizing a Hausdorff distance based approach,without adding any ex-tra computational burden.Also,no user-de?ned parameters are required.
Consider that the features here are the convex hull based invariants( , )as described in Section3.1.Thus,instead of calculating M G(U,V),H(U ,V )is calculated,where U and V refer to the2D feature-space representation of the point-sets U and V.As far as the computation of the Hausdorff distance is concerned,both MHD and EHD are used here and their performances are compared.
https://www.doczj.com/doc/c98445987.html,plexity of matching
In this section,let us mention the computational complex-ity of our matching algorithm for a query point-set consist-ing of n points and a model point-set consisting of m points. H(U ,V )is of complexity O(mn)for point-sets of sizes m and n,and can be improved to O((m+n)log(m+n)) [36].Since the af?ne invariants for a model can be computed and stored off-line,one only needs to examine the com-putational complexity for the point-set comparison phase, given a query point-set.The worst-case complexity for ob-taining the convex hull is O(n log n)and on average it is O(n).Complexity for the computation of the af?ne invari-ants is O(n).Hence,on average,the overall complexity of the matching algorithm is O((m+n)log(m+n)).
5.Experimental results
This section discusses the results obtained for the point-set matching techniques discussed in Section4.Two real databases,one consisting of dense point-sets and the other consisting of sparse point-sets,were examined in this study to compare the matching performances of the developed af?ne invariant technique with three other popular point-set matching techniques.Here,it is noteworthy to mention that although the introduced technique can handle some degree of occlusion as illustrated later in this section,it is not suitable for applications where partial matching in a scene is desired. Also,the cardinality of the point-sets should be similar,but not necessarily equal.
The effectiveness of the developed technique for real-world point-set matching problems is now mentioned for two real databases.The?rst database corresponds to the?eld photographs of a population of gray whale?ukes,consist-ing of92images taken from37individual gray whales,with 2or3different image instances per individual whale.Indi-viduals in the database were manually identi?ed by expert marine mammal biologists.The white/gray patches found on the?ukes of whales are relatively unique natural mark-ings.These patch-patterns form dense point-sets and are used here for individual identi?cation.In[37],we discussed a semi-automatic patch extraction technique using the live-wire edge detection algorithm and optimal thresholding. An example of the patch extraction outcome is shown in Fig.5.The identi?cation problem here is an image retrieval one which can be described as follows:given a query image, match it against the database and rank the database images based on their degrees of similarity with the query image. Note that the query image itself is not considered as a po-tential match.As an example,as shown in Fig.6,the?rst nine matching?ukes are shown corresponding to a given
316 C.Gope,N.Kehtarnavaz /Pattern Recognition 40(2007)309–
320
Fig.5.Extracted patch from a gray whale ?uke
image.
Fig.6.First 9identi?ed matches for a query image.
query image.In this example,the ?rst and the third match are the correct matches noting that they belong to the same whale.Afterwards,a marine mammal biologist inspects the ?rst few retrieved images to locate the correct match.The aim of the photo-identi?cation system is to limit the search space to a fraction of the database size.
The matching performance of the developed technique was also compared to two widely-used moment invariants based techniques,namely Af?ne Moment Invariants (AMI)[25]and Discrete Af?ne Moments (DAM)[27,28].It should be noted that although both of these techniques are moment invariants,they are derived differently and have different properties.In both of these techniques,the Euclidean dis-tance between the feature vectors is used for matching.In addition to the af?ne moment techniques,the performance was also compared to the convex hull-based af?ne match-ing technique in [22]mentioned earlier,where a total of 10invariants for a quadruplet of convex hull vertices are used forming so called the ?rst invariant feature vector (FIFV)and the second invariant feature vector (SIFV).This technique is denoted by convex hull invariant feature vector (CH-IFV)in Tables 4–6,bearing in mind that as mentioned in Section 4.2,only 2independent invariants can be derived for 4given points.
Table 4
Position of ?rst correct hit for various percentiles of query images Matching technique
Percentile of query images 25
5075100CH-IFV 8274689AMI 263059DAM 3113169CH-MHD 2122657CH-EHD
2
10
22
51
Comparison of the matching performances for the gray whale identi?cation problem has been shown in Table 4.The reader is reminded that the goal is not to obtain the correct match as the very ?rst hit,but to reduce the search space by having the correct match in the ?rst few retrieved images,for most (typically 75percentile)of the queries.As shown in Table 4,75%of the queries had their correct match in the ?rst 22retrieved images for the convex hull (CH)based invariants using EHD while for the AMI matching,75%of the queries had their correct match in the ?rst 30retrieved images.From the table,it can be observed that our approach was more effective in reducing the number of images to be
C.Gope,N.Kehtarnavaz /Pattern Recognition 40(2007)309–320
317
Fig.7.Two instances of the same point-pattern related by an af?ne transformation.
Table 5
Comparison of matching performance for various levels of noise Star database:Number of matching errors for 80queries (no occlusion)Noise variance Discrete af?ne moments Convex hull:MHD Convex hull:EHD CH-IFV 000001100126115512751071513111210
20
18
19
18
Table 6
Comparison of matching performance for various levels of occlusion
Star database:Number of matching errors for 80queries (noise variance =2)Occlusion %Discrete af?ne moments Convex hull:MHD Convex hull:EHD CH-IFV 04115210218512551410
28
19
20
25
searched,with EHD performing slightly better than MHD.Occlusion and noise were naturally present in the database images and hence they were not added separately.
The second studied point-set database consisted of star points,speci?ed by their 2D coordinates,corresponding to a sparse point-set situation.This point-set was used in [8]for similarity invariant point-set matching.To study af?ne invariant matching,we constructed a database consisting of 80images with 20classes,each class having four different instances of a given point-set.Each instance of a point-set consisted of 50points.Within a class,each instance was an arbitrary af?ne transformed version of every other instance.Two instances of point-sets belonging to the same class are shown in Fig.7.The matching problem studied here was,given a query point-set,?nd the best matching point-set from the database.Note that the query itself was not regarded as a potential match.If the retrieved point-set did not belong to the same class as the query,it was declared as a matching error.The matching performance was compared for various levels of noise and occlusion.Noise was introduced by having the point positions randomly perturbed via a zero mean Gaus-sian noise with various variances,as listed in Table 5.Oc-clusion was simulated by having a speci?ed percentage of the points randomly deleted from the point-sets,as listed in Table 6.As can be seen from the tables,our convex hull based matching performed signi?cantly better than the other techniques in the presence of noise as well as occlusion,
318 C.Gope,N.Kehtarnavaz/Pattern Recognition40(2007)309–320
with EHD performing slightly better than MHD.It should be noted that the performance of the AMI technique was not included in this analysis as it constituted area moments and was not applicable to discrete point-sets.
6.Conclusion
A computationally ef?cient af?ne invariant point-set matching technique has been introduced in this paper.The technique utilizes the convex hull of point-sets to obtain three af?ne invariant reference points in a novel way,which are then used to compute af?ne-invariant features.The matching is performed using Hausdorff distances in the feature space. An improved version of the MHD,named the EHD,has also been introduced and shown to generate a higher discrimina-tory power than the MHD.The developed matching tech-nique was applied to two real databases,one consisting of dense point-sets and the other consisting of sparse point-sets and the performance was compared to three other popular af?ne invariant point-set matching techniques.The results show that the developed point-set matching technique per-forms better in the presence of noise as well as occlusion. Moreover,the technique does not require any user-de?ned parameters.
7.Summary
A computationally ef?cient af?ne invariant point-set matching technique has been introduced in this paper for comparing two point-sets.The technique utilizes the convex hull of a point-set to obtain three af?ne invariant reference points in a novel way,which are then used to compute af?ne-invariant features.The matching is per-formed using Hausdorff distances in the feature space.An improved version of the MHD,named the EHD,has also been introduced and shown to generate a higher discrim-inatory power than the MHD.The developed point-set matching technique was applied to two real databases,one consisting of dense point-sets and the other consisting of sparse point-sets and the performance was compared to three other popular af?ne invariant point-set comparison techniques,namely the AMI,the DAM,and the Invari-ant Feature Vector derived from convex hulls.The results show that the technique introduced in this paper per-forms better in the presence of noise as well as occlusion. Moreover,the technique does not require any user-de?ned parameters.
Acknowledgment
This research was supported by the National Science Foundation,Grant#DBI-0077661.The gray whale?uke image database was provided by David Weller of Texas A&M University at Galveston and the National Marine Fisheries Service,Southwest Fisheries Sciences Center, La Jolla,CA.
Appendix A
Proof that area A of a polygon consisting of L vertices (x1,y1),...,(x i,y i),...,(x L,y L)is given by
A=
1
2
L
i=1
(x i y i+1?x i+1y i).
First,note that in the above summation,(x L+1,y L+1) is the same as(x1,y1),i.e.a closed polygon.The poly-gon consists of L edge segments and the i th edge seg-ment can be parameterized by a parameter t,0 t 1,as follows:
x=x i+t(x i+1?x i)?d x=(x i+1?x i)d t,
y=y i+t(y i+1?y i)?d y=(y i+1?y i)d t.(A.1) Using the Green’s theorem,we have
A=
d x d y=
1
2
(x d y?y d x).(A.2)
Let this integral for the i th edge segment be A i.Then,by substituting(A.1)into(A.2),we get
A i=
1
2
1
(x i+t(x i+1?x i))(y i+1?y i)d t
?1
2
1
(y i+t(y i+1?y i))(x i+1?x i)d t.(A.3) After simpli?cation,this area can be written as
A i=12(x i y i+1?x i+1y i).(A.4) Hence,
A=
L
i=1
A i=
1
2
L
i=1
(x i y i+1?x i+1y i).(A.5) Appendix B
Proof that centroid(c x,c y)of a polygon with an area A consisting of L vertices(x1,y1),...,(x i,y i),...,(x L,y L) is given by
c x=
1
6A
L
i=1
(x i+x i+1)(x i y i+1?x i+1y i),
c y=
1
6A
L
i=1
(y i+y i+1)(x i y i+1?x i+1y i).
C.Gope,N.Kehtarnavaz /Pattern Recognition 40(2007)309–320319
Let I denote the ?rst moment of the polygon about the x https://www.doczj.com/doc/c98445987.html,ing the Green’s theorem,we have
I = x d x d y =1
2
x 2d y .(B.1)
Using the same parameterization as in (A.1),we can evaluate (B.1)for the i th edge segment,and call it I i ,
I i =12 x 2d y
=
12 1
0(x i +t(x i +1?x i ))2(y i +1?y i )d t (B.2)After simpli?cation,this moment can be written as
I i =16(x 2i +1y i +1+x 2
i y i +1+x i x i +1y i +1
?x i x i +1y i ?x 2i +1y i
?x 2i y i ).
(B.3)
Thus,I =
L i =1
I i =16
L
i =1
(x 2i y i +1+x i x i +1y i +1?x i x i +1y i ?x 2i +1y i )
=16
L
i =1
(x i +x i +1)(x i y i +1?x i +1y i ).(B.4)
As a result,the centroid along the x -axis becomes c x =I A =16A
L
i =1(x i +x i +1)(x i y i +1?x i +1y i ),
(B.5)
where A is the area of the polygon.Similarly,c y =I A =16A
L
i =1(y i +y i +1)(x i y i +1?x i +1y i ).
(B.6)
Appendix C
Proof that centroid (c x ,c y )of a convex hull C H is af?ne invariant:
The convex polygon C H can be partitioned into N tri-angles 1, 2,..., N with areas a 1,a 2,...,a N such that a 1+a 2+···+a N =a ,where a is the area of the convex
polygon.Let (c k x ,c k y
)be the centroid of the k th triangle,for k =1,2,...,N .Thus,we can write c x =
N k =1
a k a
c k
x
,c y =
N k =1
a k a
c k y
.(C.1)
Since N k =1a k
a =1,the centroid c x ,c y can be obtained to be an af?ne combination of the centroids of the trian-gles 1, 2,..., N .Under an af?ne transformation with the transformation matrix T (assuming mean-centered points
without loss of generality),the areas of the triangles and the
convex hull are transformed as follows:
a
k
= T a k ,k =1,2,...,N
a = T a ,
(C.2)
where a k
denotes the area of the k th transformed triangle and a denotes the area of the transformed convex hull.Also,
let (p k x ,p k y
)be the transformed centroid corresponding to (c k x ,c k y ).The centroid of the transformed convex hull C H
,denoted by (c x ,c y ),is given by c
x
=
N k =1 A a k A a
p k
x
=
N k =1a k a
p k
x ,
c
y =
N k =1
A a k A a p k
y
=N k =1
a k a p k y
.(C.3)
From (C.3),we see that c x ,c y
can be expressed as the same af?ne combination of the transformed centroids and hence the centroid (c x ,c y )is af?ne invariant.
References
[1]B.N.Araabi,N.Kehtarnavaz,T.McKinney,G.Hillman,B.Würsig,A string matching computer-assisted system for dolphin photo-identi?cation,Ann.Biomed.Eng.28(2000)1269–1279.
[2]C.Gope,N.Kehtarnavaz,G.Hillman,B.Wursig,An af?ne invariant
curve matching method for photo-identi?cation of marine mammals,Pattern Recognition 38(2005)125–132.
[3]G.C.Stockman,Object recognition and localization via pose
clustering,Comput.Vision Graph.Image Process.40(3)(1987)361–387.
[4]S.H.Chang,F.H.Cheng,W.H.Hsu,G.Z.Wu,Fast algorithm for
point pattern matching:invariant to translations,rotations and scale changes,Pattern Recognition 30(2)(1997)311–320.
[5]B.Luo, E.Hancock,A uni?ed framework for alignment and
correspondence,Computer Vision and Image Understanding 92(2003)26–55.
[6]S.Umeyama,Least-squares estimation of transformation parameters
between two point-patterns,IEEE Trans.Pattern Anal.Mach.Intell.13(4)(1991)376–380.
[7]G.R.Hjaltason,M.Ray,H.Samet,I.Weiss,Using spatial sorting
and ranking in model-based object recognition,in:Proceedings of the Fourteenth International Conference on Pattern Recognition,Brisbane,August 1998,pp.1347–1349.
[8]F.Murtagh,A new approach to point pattern matching,Publications
of the Astronomical Society of the Paci?c 104(1992)301–307.[9]P.Wamelen,Z.Li,S.Iyengar,A fast algorithm for the point pattern
matching problem,IEEE Trans.Pattern Anal.Mach.Intell.,last revised Dec.1999.
[10]D.Huttenlocher,S.Ullman,Recognizing solid objects by alignment
with an image,https://www.doczj.com/doc/c98445987.html,puter Vision 5(2)(1990)195–212.
[11]H.Wolfson,I.Rigoutsos,Geometric hashing:an overview,IEEE
Comput.Sci.Eng.4(1997)10–21.
[12]D.Huttenlocher,D.Klanderman,A.Rucklige,Comparing images
using the hausdorff distance,IEEE Trans.Pattern Anal.Mach.Intell.15(9)(1993)850–863.
[13]W.Rucklidge,Ef?ciently locating objects using the hausdorff
distance,https://www.doczj.com/doc/c98445987.html,put.Vision 24(3)(1997)251–270.
[14]M.Hagedoorn,R.Veltkamp,Reliable and ef?cient pattern matching
using an af?ne invariant metric,https://www.doczj.com/doc/c98445987.html,put.Vision 21(2–3)(1999)203–225.
320 C.Gope,N.Kehtarnavaz/Pattern Recognition40(2007)309–320
[15]Z.Zhang,Iterative point matching for registration of free-form
curves,https://www.doczj.com/doc/c98445987.html,put.Vision13(2)(1994)119–152.
[16]R.Sarfari,N.Narsimhamurthi,M.Shridhar,M.Ahmadi,Af?ne
point pattern matching and its application to form registration,in: Proceedings of the IEEE Symposium on Circuits and Systems, vol.2,Atlanta,May1996,pp.656–659.
[17]K.Arbter,W.E.Snyder,H.Burkhardt,G.Hirzinger,Application
of af?ne-invariant fourier descriptors to recognition of3-D objects, IEEE Trans.Pattern Anal.Mach.Intell.12(7)(1990)640–647. [18]https://www.doczj.com/doc/c98445987.html,mdan,J.T.Schwartz,H.J.Wolfson,Af?ne invariant model-
based object recognition,IEEE Trans.Robot.Automat.6(5)(1990) 578–589.
[19]M.S.Costa,R.M.Haralick,L.G.Shapiro,Optimal af?ne invariant
pattern matching,in:Proceedings of the Tenth International Conference on Pattern Recognition,vol.1,1990,pp.233–236. [20]T.Wakahara,Y.Kimura,A.Tomono,Af?ne-invariant recognition of
gray-scale characters using global af?ne transformation correlation, IEEE Trans.Pattern Anal.Mach.Intell.23(4)(2001)384–395. [21]A.Goshtasby,G.C.Stockman,Point pattern matching using convex
hull edges,IEEE Trans.Syst.Man Cybern.SMC-15(1985) 631–637.
[22]Z.Yang,F.S.Cohen,Image registration and object recognition using
af?ne invariants and convex hulls,IEEE Trans.Image Processing8
(7)(1999)934–946.
[23]D.Huttenlocher,Fast af?ne point matching:an output-sensitive
method,in:Proceedings of the Computer Vision Pattern Recognition, Maui,June1991,pp.263–268.
[24]Z.Yang,F.Cohen,Cross-weighted moments and af?ne invariants for
image registration and matching,IEEE Trans.Pattern Anal.Mach.
Intell.21(8)(1999)804–814.
[25]J.Flusser,T.Suk,Af?ne moment invariants:a new tool for character
recognition,Pattern Recognition Lett.15(1994)433–436.
[26]T.Reiss,The revised fundamental theorem of moment invariants,
IEEE Trans.Pattern Anal.Mach.Intell.13(8)(1991)830–834.[27]H.Suesse,W.Ortmann,Robust matching of af?nely transformed
objects,in:Proceedings of the International Conference on Image Processing,Barcelona,September2003,pp.II-375–378.
[28]K.V oss,H.Suesse,Af?ne point pattern matching,in:Pattern
Recognition,Lecture Notes in Computer Science,vol.2191,Springer, Berlin,2001,pp.155–162.
[29]J.Heikkil?,Pattern matching with af?ne moment descriptors,Pattern
Recognition37(9)(2004)1825–1834.
[30]J.Sprinzak,M.Werman,Af?ne point matching,Pattern Recognition
Lett.15(1994)337–339.
[31]H.Coxeter,Introduction to Geometry,Wiley,New York,1969,pp.
216–221.
[32]C.Barber,D.Dobkin,H.Huhdanpaa,The quickhull algorithm for
convex hulls,ACM Trans.Math.Softw.22(4)(1996)469–483. [33]F.Preparata,M.Shamos,Computational Geometry:An Introduction,
Springer,Berlin,Germany,1985.
[34]P.F.Stiller,General approaches to recognizing geometric
con?gurations from a single view,Proc.SPIE:Vision Geom.VI 3168(1997)262–273.
[35]M.Dubuisson,A.Jain,A modi?ed hausdorff distance for object
matching,in:Proceedings of the Twelfth Internatinal Conference on Pattern Recognition,Jerusalem,Israel,October1994,pp.566–568.
[36]H.Alt, B.Behrends,J.Blomer,Measuring the resemblance of
polygonal shapes,Proceedings of the Seventh ACM Symposium on Computer Geometry(1991).
[37]N.Kehtarnavaz,V.Peddigari, C.Gope,W.Syed,G.Hillman,
B.Wursig,Photo-identi?cation of humpback and gray whales
using af?ne moment invariants,in:Proceedings of the Thirteenth Scandinavian Conference on Image Analysis,Sweden,June–July 2003.
[38]M.P.Yu,K.C.Lo,Object recognition by combining viewpoint
invariant fourier descriptor and convex hull,in:Proceedings of the International Symposium on Intelligent Multimedia,Video and Speech Processing,Hong Kong,May2001,pp.401–404.
About the Author—C.GOPE received the B.Tech.degree from Indian Institute of Technology,Kharagpur in Electrical Engineering in2000and the M.S.degree from Washington State University in Electrical Engineering in2002.He is currently pursuing a Ph.D.degree in Electrical Engineering at the University of Texas at Dallas.His research interests include image processing,pattern recognition,and computer vision.
About the Author—N.KEHTARNA V AZ received the Ph.D.degree in Electrical and Computer Engineering from Rice University in1987.He is currently a Professor of Electrical Engineering at the University of Texas at Dallas.Previously he was a Professor of Electrical Engineering at Texas A&M University.His research interests include signal and image processing,pattern recognition,and real-time imaging.He has authored or co-authored?ve books and more than130journal and conference papers in these areas.He is currently serving as Co-Editor-in-Chief of Journal of Real-Time Image Processing,and Chair of the Dallas Chapter of the IEEE Signal Processing Society.Dr.Kehtarnavaz is a Fellow of SPIE,a Senior Member of IEEE and a Registered Professional Engineer.More information on Dr.Kehtarnavaz’s research activities are available at https://www.doczj.com/doc/c98445987.html,/～kehtar
校园景色（1单元） 春姑娘悄悄地来到我们身边，冰雪融化，万物复苏，那青的草，绿的叶，各种色彩鲜艳的花,给校园构成一幅生机勃勃的春景图。 在校园的草坪上，可以看见嫩绿的小草从土壤里探出头来，高兴地望着我们，仿佛在说：“亲爱的同学们，我们又见面了！”一个个圆圆的蘑菇，在春雨的浇灌下，也露出了笑脸。 校园的主道两旁，杨树挺立在那里，一动不动，像一位站岗的哨兵。枝条上长出了圆圆的新叶，远远望去，像一个绿色的小桃子。那一朵朵美丽鲜艳的花儿，争齐开放，使人感到格外舒服。 每当下课的时候，悦耳的铃声传来，同学们就像一只只鸟儿一样“飞”出了教室，校园里顿时一片欢腾。兵乓球台前，围满了人，一阵阵叫好声，一阵阵欢呼声，使校园里充满了活力，充满了欢乐。 美丽的春天，美丽的校园，真是一幅美丽的画卷。 校园春色（1单元） 春姑娘悄悄地来到我们身边，冰雪融化，万物复苏，那青的草，绿的叶，各种色彩鲜艳的花，给校园构成一幅生机勃勃的春景图。 在校园的草坪上，可以看见嫩绿的小草从土壤里探出头来，高兴地望着我们，仿佛在说：“亲爱的同学们，我们又见面了！”一个个圆圆的蘑菇，在春雨的浇灌下，也露出了笑脸。 校园的主道两旁，杨树挺立在那里，一动不动，像一位站岗的哨兵。枝条上长出了圆圆的新叶，远远望去，像一个绿色的小桃子。那一朵朵美丽鲜艳的花儿，争齐开放，使人感到格外舒服。 每当下课的时候，悦耳的铃声传来，同学们就像一只只鸟儿一样“飞”出了教室，校园里顿时一片欢腾。兵乓球台前，围满了人，一阵阵叫好声，一阵阵欢呼声，使校园里充满了活力，充满了欢乐。 美丽的春天，美丽的校园，真是一幅美丽的画卷。 爸爸，我想对你说 爸爸，我想对你说：你的父爱我能感受到，但你能不能不要太严了呢？ 每天，我都要写一篇日记，但有时候作业太多，你还硬逼着我写日记；你不知道，我每次写完作业都很累很累，好想看一下电视，放松一下，但你硬逼我写，没办法，我只好受点委屈了。其实，我有时很不想学，但你又非要我学不可，我又不能不听你的话，只好跟你学，你不知道，我不想学的时候，你讲的知识，我一点也没听进去，所以你等于白讲了。你还把我心爱的漫画书藏起来，买来一大堆的试卷，叫我学到哪课就做到哪课，那里我是多么想看漫画书啊！有时，我放学回家把电视机打开看，但你却叫我过去写作业，那时我是多么不乐意呀！因为你不让我看电视，而你却在那里看电视。你知道吗？我人虽然在这边，但心却飞到电视机那里去了，所以作业才会出现一些不该错的地方。每天晚上，我叫你讲故事给我听，但你却拒绝了，叫我自己看，但睡着看书容易损伤眼睛，所以又看不成，你不知道，我那时是多么想听你讲那动听的故事呀！ 爸爸，你说过，我们既是父子也是朋友，那你能不能给我一个自由的天地，让我自由地发展呀！
《美国历史与文化》 结课论文 专业：化学工程与工艺 学号：041114116 姓名：杨乐
Columbo's influence on the American continent Columbo is a famous Spanish navigator, is a pioneer of the great geographical discovery. Columbus young is garden said believers, he so haunting had in Genoa made prison of Marco Polo, determined to be a navigator.1502 he crossed the Atlantic four times in 1492, discovered the American continent, he also became a famous navigator. On August 3, 1492, Columbus by the king of Spain dispatch, with credentials of Indian monarchs and emperors of China, led the three tons of Baishi of sailing, from Spain Palos Yang Fan of the Atlantic, straight towards to the West. After seventy days and nights of hard sailing, in the early morning of October 12, 1492 finally found the land. Columbo thought he had arrived in India. Later know that Columbus landed on a piece of land belonging to the now Balak America than the sea in the Bahamas, he was it named San Salvador. Columbo's discovery has changed the course of world history. It opens up a new era of development and colonization in the new world. At that time, the European population was expanding, with this discovery, the Europeans have settled in two new continents, there will be able to make a difference in the European economy and the resources of the mineral resources and raw materials. This discovery led to the destruction of the American Indian civilization. From a long-term point of view, there have been a number of new countries in the Western
【篇一】高中人与自然作文800字 大自然是奇妙无比的，它蕴含着许许多多不为人知的秘密，等待我们去发现，去探索。好奇心驱使着人们全身心的投入到了对大自然的研究当中，得出了很多有价值的结论，并把它们记录下来，编成了一部部对后人很有影响的书籍。《昆虫记》就是其中最富盛名的代表作。 《昆虫记》这本书主要记载了法国生物学家法布尔四十多年来不断研究昆虫所得出的结论与体会。书中的那些昆虫们栩栩如生，让读者好像亲眼见到了他们一样。而且其中大部分昆虫都是鲜为人知的，更加的富有新意，让人读完以后回味无穷，实在不失为一部佳作哪！ 我们不得不佩服法布尔先生妙笔生花的才华。《昆虫记》不单单是平铺直叙的介绍昆虫，而且是以“我”为人称，用叙事体的手法，以一位观察者的角度来写的，又不缺少亲切感，使读者能够能加侧面的了解他们。 法布尔在介绍每一种小昆虫的时候，首先先介绍他们的肢体、颜色、花纹、形状、特征等外部构造，更好地使读者认识他们，并为下文打好了基础。然后，把他们的生活习性做一简单的阐述。最后，大篇幅的介绍他们捕食、交配、繁殖等具有代表性的几个方面，并在其中穿插记叙了自己在观察研究中所发生的趣事，更加让文章生动有趣，具有吸引力。 读完这本书以后，我思绪万千，受益匪浅。不仅仅是丰富了自己的阅历，而更多的则是对小昆虫的刮目相看。以前，我对那些不起眼的小昆虫不屑一顾，现在则对他们有了一个全新的认识。就拿蚂蚁来说吧，他们起早贪黑，兢兢业业；他们团结一心，通力合作；他们巧用智慧，机灵伶俐；他们不畏艰辛，全力以赴；他们互帮互助，团结友爱；他们懂得分享，懂得谦让；他们诚实守信，不予欺骗；他们舍己为人，以小顾大；他们分工明细，各尽其职……微小的蚂蚁，竟有如此崇高的品质，顽强的精神，我被深深的震撼了。是啊，连人类都望尘莫及，无法做到完全具备的，蚂蚁却做到了。所以，我们不能小看每一个生物，人不可貌相，海水不可斗量啊！ 大自然就是这样的丰富多彩。读了《昆虫记》，让我更加懂得爱护昆虫，兴致勃勃的欣赏有趣的昆虫。他们实在是太有趣了！致力于此类事业，真是不亦乐乎那！ 【篇二】高中人与自然作文800字 从教科书到影视，人与自然这个主题的作品有很多，最近这段时间，学会了一些课文，偶然看一两部电影，恰巧都是这个主题，引起了我的深思。 严春友的《敬畏自然》，一开篇就说明了人类征服自然的想法有多可笑，“不自量力地要用这滴水来代替大海”，而事实上，人类确实也这样做了，“毁林开荒”“围湖造田”就是很典型的例子，以为可以用人力改变自然，可结果呢？由
高考语文作文备考：名校模拟题作文及例文汇编 四、写作（60分） 10．阅读下面的材料，根据要求写作。 近日，某中学举行了一场以“拥有了实力，行事是该高调还是该低调”为主题的辩论赛。 正方的观点是：拥有了实力，行事就应高调。因为拥有了实力，你也掩盖不了，高调做事是一种责任，一种气魄，一种精益求精的风格，一种执著追求的精神。 反方的观点是：拥有了实力，行事也应低调。因为做事必先做人，低调是一种姿态，一种修养，一种风度，一种智慧，一种谋略。 时代发展到今天，无论个人、集体还是国家，你觉得拥有了实力应该采取何种态度呢？请你在正反双方中选择一方观点作为自己的立场，写一篇辩论稿。要求：选好角度，确定立意，自拟标题，注意格式；不要套作，不得抄袭，不少于800字。 【分析】本题考查学生写作的能力。本文属于任务驱动型文章。写作时要整体理解把握材料的内容，明白材料蕴含的道理，选取合适的角度进行立意构思。写作指令：选择辩题，写一篇辩论稿。写好材料作文的关键在于理解材料主旨并进行立意。本题材料的中心事件是某中学举行的一场以“拥有了实力，行事是该高调还是该低调”为主题的辩论赛，正方的观点是拥有了实力，行事就应高调，反方的观点是拥有了实力，行事也应低调。考生可以选择正方观点，指出高调做事是一种责任，一种气魄，一种精益求精的风格，一种执著追求的精神；可以选择反方观点，指出做事必先做人，低调是一种姿态，一种修养，一种风度，一种智慧，一种谋略。最后还要分析文体，材料要求写成“辩论稿”，这就要求观点要明确，辩论要有力，论据要充足。无论选择哪一个辩题，分析都要有思辩性，切不可完全否定对方的观点（将对方说得一无是处）。 参考立意：1、高调做事是一种责任，一种气魄。2、低调做人是一种修养，一种风度。 【解答】低调做人是一种智慧 尊敬的各位评委老师、同学们： 大家好，我方的辩题是“低调做人是一种智慧”。 山不解释自己的高度，并不影响它的耸立云端；海不解释自己的深度，并不影响它容纳百川；地不解释自己的厚度，但没有谁能取代她作为万物的地位……做人，我认为关键是要学会低调，低调是一种人生的智慧。 莫言是中国第一个获得诺贝尔奖的人，圆了所有中国人几个世纪的“诺奖梦”。但让人可歌
浅谈美国历史 ——见证从蚂蚁到大象历程 引导语：中华文化源远流长，五千年的华夏文明留给我们太多的回忆。分久必合、合久必分；从繁荣昌盛到民族衰落；从压迫受辱到当家作主、从璀璨奇葩到复兴中华……可是，跨过大洋的彼岸，初出茅庐的美国却在近两百多年的历史跨度下完成了从蚂蚁到大象的历程，创造出美国独特的发展文化。今天的世界，“汤姆大叔”在全球“维护着世界和平”；好莱坞大片充斥着各大荧屏；NBA回荡在茶前饭后的娱乐中……两百多年来，美国历史一直都是民主制度的试验。早年被提出的问题如今持续被提出并获得解决；强大政府对抗弱小政府、个人权利对抗群体权利、自有资本主义对抗受到管理的商业与劳工以及参与世界对抗孤独主义。美国对于民主制度有很高的期待，而现实又是不如人意。然而国家经过妥协，已见成长与繁荣。在今天的发展过程中笔者认为有必要借鉴美国蚂蚁变大象的历程。 自从哥伦布发现新大陆之后，这片土地上开始了她不平凡的发展。十七世纪初，英国开始向北美殖民。最初的北美移民主要是一些失去土地的农民，生活艰难的工人以及受宗教迫害的清教徒。在殖民地时代，伴随着与北美洲原住民印第安人的长期战争，对当地印第安人的肆意屠杀，严重的劳力缺乏产生了像奴隶和契约奴隶这类的非自由劳力。万恶的黑奴贸易盛行起来。从1607年到1733年，英国殖民者先后在北美洲东岸建立了十三个殖民地。由于英国移民北美是为了追求自由和财富，如被迫害的清教徒和贫农。地方政府享受自治权。殖民地居民有比英人更广泛参与政治的机会和权利，培养了自治的意识和能力，所以他们相信社会契约中，政府是人民需要保护而得人民支持才组成的。在十八世纪中期,殖民地的经济，文化，政治相对成熟，殖民地议会仍信奉英王乔治三世，不过他们追求与英国国会同等的地位，并不想成为英国的次等公民，但是此时英法的七年战争结束，急于巩固领土，使向北美殖民地人民征租重税及英王乔治三世一改放任政策，主张高压手段。因此引发殖民地人民反抗，如“没有代表就不纳税”宣言、“波士顿惨案”、“不可容忍的法案”等。1775年4月在来克星顿和康科特打响“来克星顿的枪声”揭开美国独立战争的前奏。后来，这些殖民地便成为美国北美独立十三州最初的十三个州。 1774年, 来自12州的代表，聚集在费城, 召开所谓第一次大陆会议，希望能寻出一条合理的途径, 与英国和平解决问题，然而英王却坚持殖民地必须无条件臣服于英国国王, 并接受处分。 1775年，在麻州点燃战火, 5月，召开第二次大陆会议, 坚定了战争与独立的决心，并起草有名的独立宣言, 提出充分的理由来打这场仗，这也是最后致胜的要素. 1776年7月4日，宣告了美国的独立，1776年7月4日大陆会议在费城乔治·华盛顿发表了《独立宣言》。《独立宣言》开宗明义地阐明，一切人生而平等，具有追求幸福与自由的天赋权利；淋漓尽致地历数了英国殖民主义者在美洲大陆犯下的罪行；最后庄严宣告美利坚合众国脱离英国而独立。《独立宣言》是具有世界历史意义的伟大文献。完全脱离英国，目的是为‘图生存、求自由、谋幸福’，实现启蒙运动的理想。 1781年, 美军赢得了决定性的胜利， 1783年, 美英签订巴黎条约，结束了独立战争。这也充分展现出
电脑绘画班WINDOWS画图详细教程 目录 一.如何使用画图工具 二.《画图》工具系列-妙用曲线工具 三. 《画图》工具系列-巧用圆形工具 四. 《画图》工具系列妙用文字工具 五. 用“画图”进行屏幕拷贝 六. “画图”程序的放大修改功能 七. “画图”中的工具与颜色配置 八. 灵活使用编辑功能 九. Windows画图程序操作技巧 十. Windows画图程序操作技巧 十一. 用画图程序检测LCD的暗点
一.如何使用画图工具 想在电脑上画画吗？很简单，Windows 已经给你设计了一个简洁好用的画图工具，它在开始菜单的程序项里的附件中，名字就叫做“画图”。 启动它后，屏幕右边的这一大块白色就是你的画布了。左边是工具箱，下面是颜色板。 现在的画布已经超过了屏幕的可显示范围，如果你觉得它太大了，那么可以用鼠标拖曳角落的小方块，就可以改变大小了。 首先在工具箱中选中铅笔，然后在画布上拖曳鼠标，就可以画出线条了，还可以在颜色板上选择其它颜色画图，鼠标左键选择的是前景色，右键选择的是背景色，在画图的时候，左键拖曳画出的就是前景色，右键画的是背景色。 选择刷子工具，它不像铅笔只有一种粗细，而是可以选择笔尖的大小和形状，在这里单击任意一种笔尖，画出的线条就和原来不一样了。
图画错了就需要修改，这时可以使用橡皮工具。橡皮工具选定后，可以用左键或右键进行擦除，这两种擦除方法适用于不同的情况。左键擦除是把画面上的图像擦除，并用背景色填充经过的区域。试验一下就知道了，我们先用蓝色画上一些线条，再用红色画一些，然后选择橡皮，让前景色是黑色，背景色是白色，然后在线条上用左键拖曳，可以看见经过的区域变成了白色。现在把背景色变成绿色，再用左键擦除，可以看到擦过的区域变成绿色了。 现在我们看看右键擦除：将前景色变成蓝色，背景色还是绿色，在画面的蓝色线条和红色线条上用鼠标右键拖曳，可以看见蓝色的线条被替换成了绿色，而红色线条没有变化。这表示，右键擦除可以只擦除指定的颜色－－就是所选定的前景色，而对其它的颜色没有影响。这就是橡皮的分色擦除功能。 再来看看其它画图工具。 是“用颜料填充”，就是把一个封闭区域内都填上颜色。 是喷枪，它画出的是一些烟雾状的细点，可以用来画云或烟等。 是文字工具，在画面上拖曳出写字的范围，就可以输入文字了，而且还可以选择字体和字号。 是直线工具，用鼠标拖曳可以画出直线。 是曲线工具，它的用法是先拖曳画出一条线段，然后再在线段上拖曳，可以把线段上从拖曳的起点向一个方向弯曲，然后再拖曳另一处，可以反向弯曲，两次弯曲后曲线就确定了。 是矩形工具，是多边形工具，是椭圆工具，是圆角矩形，多边形工具的用法是先拖曳一条线段，然后就可以在画面任意处单击，画笔会自动将单击点连接
写印象深刻的人的作文800字_高一作文 写印象深刻人作文800字 小胡子青年 去年夏天，我独自登上了去Q市轮船。 没想到四等舱也闷热，更晦气是与这样一对“宝贝”做“舱友”——那男青年留着小胡子，时不时扯起破嗓门儿大唱流行歌曲，唱得人一阵阵起鸡皮疙瘩；他女朋友一张娃娃脸，不是描眉扑粉，就是吃零食。唉，没素质！ 终于到了避暑胜地Q市一踏上码头，我长长舒了口气，心想：回上海时可千万别再碰上这样“宝贝”了！ 我爱游泳，一到婶婶家先打听海水浴场在哪里，第二天便兴冲冲赶去。换上泳装，跳进大海，沐浴着凉爽海水，舒服极了。 以后我天天奔海水浴场。徜徉在海阔天空之间，像鱼儿一样游来游去，第一次体会到了“不管风吹浪打，胜似闲庭信步”快乐。 一天，婶婶一家带我去游L山，下午回来后我看天色还早，便又去了海水浴场。那天游客不多，我游得自由自在，不知不觉离沙滩远了。义游了一会儿，渐渐感到海水晃得厉害，猛然想起，莫非退潮了？ 我赶紧往沙滩游，可是体力不支。我拼命游呀游，突然，脚心一凉，啊，脚抽筋了！这时候偏偏一个浪头打来，我来不及换气，“咕噜咕噜”喝了几口海水，便身不南己地往下沉…… 不知过了多久，听见四周有人在说：“好了好了，她醒了！”“唉，好险！”……又觉得两只手臂被人抓着摆过来摆过去。啊，难道我刚才差点儿淹死了？后来又被人救了？真可怕！这时候浑身软绵绵，嘴巴发苦，努力睁开眼睛，迷迷糊糊看见一个小青年在给我做人工呼吸。再一看，他好生面熟。噢，想起来了，是同船来那个小胡子青年！ “喂，小朋友，你不会游泳别逞能。今天要不是碰上我，你就回不了上海啦！”他放下我手臂，站起来，搓搓手，挽起女朋友，一扭一扭地走了。 我想说感谢他话，可是没有力气。望着他俩远去背影，心里又感激又惭愧。 听着海浪声，忽然想起了这样一句俗语：“人不可貌相，海水不可斗量。”是呀，事物是复杂，人似乎更复杂。我怎么可以戴一副有色眼镜去认识社会上形形色色人呢？ 点评：
2018高考任务驱动型作文习题及范文汇编 阅读下面的材料，根据要求写一篇不少于800字的文章。（60分）【原创预测题8】《南方周末》报道，2016年4月14日，在山东发生了一起辱母杀人案。女企业家苏银 霞欠了当地涉黑团伙头目吴学占的高利贷，仍欠17万未还清。被于欢刺死的杜志浩等人有 涉黑行为，辱骂殴打、限制并侮辱于母等情节恶劣，警察调解无效后，于欢拿起办公室桌上水果刀，乱刺催债人员，至一死三伤,, 2017年3月17日山东省淄博城市法院一审以故意伤害罪判处于欢无期徒刑，还需赔偿死者家属30598.5元和伤者53443.47元。在社会引起强烈深刻的反响。案件进展：全国高 检派人、山东高法介入指导工作。2017年6月23日山东高法二审裁定于欢犯故意伤害罪， 判处有期徒刑五年并经济赔偿。 比较情、理和法三者，你认为哪个更重要？请综合材料内容及含意作文。要求选好角度，确定立意，明确文体，自拟标题；不要套作，不得抄袭。 法律，高于情理的利剑 ①2016年，在山东发生一起辱母杀人案。由于嫌疑人于欢杀人合情合理，且受害者为 涉黑人员，引起社会强烈反响。最终涉黑者被抓，于欢二审得到减刑。此案看似于欢为母出气，合情合理但不合法。法律是高悬情与理之上的利剑，它约束情与理，以免失控。 ②从事件本身看，我认为，于欢因母亲受辱，在多方求助无果的情况下挥刀，这个举动 似乎合情合理。哪一个孩子看见母亲受辱能无动于衷？但用刀伤人，属于防卫过当，造成严重后果就违法了，他被判刑五年，这是对冲动的惩罚。正因为于欢所为合情合理，因而才引
发社会强烈反响，减轻刑罚并惩处恶人。但合情合理不能作为评判的唯一标准，法律是不可逾越的红线，它是约束情与理的尺度。情理不能大于法，否则社会就会混乱。 ③若没有法律约束，人便为所欲为。荀子曰“木受绳则直，金就砺则利。”进入依法治国时代，任何事物都必须按章办事，一视同仁。就拿网购说吧，它是一件新鲜事，存在利润 空间。初期因法律不健全，假货次品充斥市场，诚信遭遇冰河，网购受到巨大损失。当相关 法律出台，市场秩序迅速恢复，网络经济成为我国支柱产业。因此，合情合理又守法，就会 得到保护，避免遭受损失。最近出台的“信联”措施就很好，严惩那些不讲诚信的违法者。 ④合情合理是谁都想要的但必须依法办事，以其道得之，最终不受处罚。其实，想要减少情与理的影响，依法办事也容易。首先，应明法懂法守法。懂法的途径是读法律条文；其 次，遇事时要冷静，不要完全被情理支配。当然了，法律建立在情理之上，在法律范围内， 保护合情合理的事情。本次事件，于欢正被情理冲昏头脑忘却法律，最终导致悲剧发生。 ⑤因此，我认为在处理看似合情合理的事物前，要理性与理智的思考，自己行为的利弊得失，然后在行动。最后，请时刻谨记康德的话：世界上有两样东西是我最敬畏的，一是天 上的星星，另一个是道德法律。知法明理，敬法守法。只有如此，才不会完全被情理支配， 依法办事，就不会出现问题。 ⑥然而，守法并不意味忘却情理。在现实社会中，经常出现“扶不扶”“救不救”的两难事情。试问那些不扶的人违法吗？没有。他们“不为”却践踏道德底线，伤害情理。那么 世界就会变得冷漠。好在新的法律保护讲道德见义勇者，他们要受到奖励和表彰。 ⑦法律是高悬于情理之上的利剑，遵法守纪是情理的红线。铭记教训，莫让悲剧重演。 辽宁省沈阳市第11中学2018届17班李浩然指导教师：孙延堂
宗教对美国社会的影响 梁子毓英语学院英语137班 摘要：美国是当今世界上最发达的资本主义国家，同时又是发达国家中最具宗教色彩的国家，可以说美国的发展与宗教有着密不可分的关系。宗教从美国建国伊始至今对人们思想的影响是根本性的，这种影响绝不仅仅只是停留在道德方面，在政体的确立、民主制度的促进等方面都有着同样深远的作用。如今，宗教在美国的影响力有增无减，信仰上帝的人越来越多了，宗教在美国国家社会生活中的作用也越来越大。因此，研究宗教对美国发展的影响更有助于我们了解美国的社会现实和文化特征，可以使我们对宗教在美国社会的影响有更深刻的认识。 关键词：宗教美国社会影响 引言 美国文化的一大特征就是其宗教性。来自世界各地不同国家不同种族的移民带来了他们自己的宗教，使美国成为多宗教的国家。各种不同的宗教必然会对美国的社会产生影响，同样也融入了美国的文化。在众多宗教中，影响最大的是新教众多教派中的清教，清教在发展过程中，其影响超越了宗教领域，渐渐渗透到了美国宗教以外的政治、经济、文化领域，使美国政治、经济、文化都带有明显的清教主义特征。清教主义因而成为美国文化和社会形成过程中的重要因素。美国的文明都刻有明显的清教主义印记，清教主义文化也造就了独特的美利坚精神与文化。 一、基督教新教对美国历史的影响 美国虽然只有四百多年的历史，但却从一个英属殖民地逐渐成为世界上的头号强国。在这片土地上诞生了世界上第一部成文宪法，产生了世界上第一个民选总统，建立了三权分立的民主政体，这些都对世界历史产生了深远的影响。然而这些都与基督新教及其伦理道德有着密不可分的关系，这种关系甚至是决定性的。 早期殖民北美的新教1徒的新教信仰，构成了北美早期社会思想风潮的主调，也构成了以后美利坚的民族精神以及国家意识形态的基础。美国独立战争的发生，是因为早期移民北美的多数人都是新教教徒的缘故。由逃亡的清教徒们建立的美洲殖民地，在宗教上，一直与英国本土的宗教处于对立状态。美洲大陆的宗教主流为清教徒，英国的国教则为天主教与新教的混合体圣公会安力甘宗，安力甘宗作为英国国教一直是清教徒改革的对象。在美国独立战争及18世纪二十年代，英国本土和美洲殖民地发生了一场轰轰烈烈的宗教“伟大复兴”运动，这场运动在美国是新教教义的普及和强化运动，是一场彻头彻尾的新教教义在新大陆被强化的运动，这场运动最后导致了新教的进一步振兴，从而在思想上与英国国教彻底脱离了关系。宗教运动进一步促进了北美殖民地人群的主体意识，进一步加强了殖民地与英国本土的在宗教上和政治上的离心力，为独立战争做了思想和意识形态的准备。北美殖民地与英国本土上的宗教对立，对美国独立战争的爆发有着深远的影响。 对美国近代发展奠定基础的南北战争，实际上也有着深刻的宗教原因——美国清教对南方奴隶制的憎恶而引起的南北方对立。废奴主义一直是基督教教义的传统。在新教中这被理解为上帝爱世上的每一个人。这种思想成为西方人权思想的源头。既然上帝爱每一个人，个 1新教：新教（Protestantism）是由16世纪宗教改革运动中脱离罗马天主教的一系列新教派的统称，主流教派有路德宗、加尔文宗和圣公会。
每个人都有自己擅长的写作类型，不断练习相应作文的过程中，我们的作文会逐渐形成自己独有的写作习惯。以下是愛师网小编为您整理的各有绝活作文身边的绝活作文别人的绝活作文。希望您能喜欢。 各有绝活作文身边的绝活作文别人的绝活作文第一篇： 爸爸的绝活 我的爸爸有一项绝活，炒菜，就是这个绝活，让我十分的佩服我爸爸，下面我就跟大家说一说吧。 我的爸爸是个矮个子，他长着浓眉大眼，高高的鼻子，那棱角分明的嘴唇，平添了几分英俊和自信。 爸爸大部分都是妈妈帮他盛饭，把菜夹到饭里，爸爸再把饭端到电脑室，一边吃饭一边玩游戏或看电视。他有一张“血盆大口”每次吃饭都是“狼吞虎咽”而且一口就能喝下半杯水。吃完饭后又是我帮爸爸把碗端出去给妈妈洗。这天也不例外。真想让爸爸改了这个坏毛病。 但是，爸爸偶尔也会给我们炒一顿好菜。爸爸以前学过厨艺。他切菜动作很灵活。切菜时，膀子不动手腕动，速度很快。不像妈妈，拿刀时膀子整个动，好像使出九牛二虎之力，切完菜她直喊累。爸爸切出的菜都是一样粗细。妈妈切出的菜却是大大小小的，不如爸爸。爸爸炒菜时喜欢用右手拿着锅铲翻菜，不时地还用左手拿着锅把子一抖，我真怕菜掉外面烫到爸爸的手。只见菜从锅中直向上冲，然后翻了个身，又听话地回到锅中。看来我的担心是剩余的。 这就是我爸爸最厉害的绝活，是不是十分的厉害呢，但是我爸爸平时不显山不露水的，想要看到这项绝活还找呢不容易。 各有绝活作文身边的绝活作文别人的绝活作文第二篇： 绝活作文 “六一”儿童节快到了，妈妈说要送我一双旱冰鞋。那天，鞋店我挑了许多鞋，最后我选中了一双灰色的，因为我觉得男孩子还是灰色大方一点。 回家后，我急忙拿出鞋子，想试穿一下。穿进去刚想站起来，我就滑了一跤。我用双手扶住墙慢吞吞地爬起来。我又试图往前走几步，没想到我的脚自动向前滑，我嘴巴里“哎呀”还没说完，就摔了个四脚朝天。我疼得直想哭，但我硬得撑住，我可不能让妈妈小看我。我一边椅子上休息，一边脑子里想着不能放下，该想个什么法子。 过一会儿，我又开始练习了，这次，我草坪上先练走步，双手拿了两根木棒，并带上护膝，以防万一。果然这样练习，比较保险，我能自如地草地上走了，我心里甭提多高兴了。
中考作文范文汇编10篇 一、作文写作 1．阅读下面的文字，按要求完成微作文。 学校准备为高一新生设计制作校服。原定为一款参照国外校服样式设计的校服，男式帅气，女式俏丽，大家很喜欢。但有家长提出，这样的校服穿上后容易使学生分心，甚至会助长男女同学之间的爱慕之心。校方对此很为难。 请你给校长写一封短信，陈述自己的看法，对解决这一问题提供帮助。要求表达得体，150字左右。 【答案】示例：尊敬的校长： 关于高一新生的校服，我们希望学校能采用原定的设计方案。因为这款校服帅气美丽，能尽显青春气息，展示我们的阳光与朝气。至于有些顾虑，我们可以理解，但并不赞同。因为校服是每个人都穿，大家都美，天天都穿，很是平常。如果某一事物完全融入到我们的生活之中时，我们就不会因此而生出什么特别的关注，只有生活会因此多出一份美丽。我们期待着这份美丽。 祝您工作顺利，心情愉快！ 高一学生：×××【解析】 这是一篇小作文。首先要仔细分析给出的材料，了解到学校定制的这套参照国外校服样式设计的校服深受学生的喜爱，可是学校方面有顾虑。给校长的信要申明学生的观点，观点要具体明确，符合题意，见解深刻，言之有理，有助于解决问题。注意字数的要求。 点睛：这是一篇微作文。“微写作”发源于“微博”“微信”，一般是指根据一定的情景或要求，写作200字左右的“微作文”。“微写作”对考生的写作能力和综合素养提出了更高的要求，中考“微写作”要求考生文字的表达、意境的创造等诸多方面，均应超过网络“微写作”。完成这篇微作文，要明确情境的内容，选择合适的角度写作。 2．作文。 “一切都将会过去；而那过去了的，就会成为亲切的怀恋。”这是普希金《假如生活欺骗了你》中的诗句。诗人告诉我们：生活不可能一帆风顺，当越过坎坷，蓦然回首时，你往往会发现，那些过去了的都已成为人生宝贵的财富，成为你“亲切的怀恋”。同学们，你也曾有过这样的经历和感受吧？请自拟题目，写一篇文章与大家分享。 要求：(1)题目自拟：(2)紧扣材料主题，内容具体充实：(3)有真情实感：(4)文体不限（诗歌、戏剧除外）；(5)不少于600字；(6)文中请回避与你相关的人名、校名、地名。【答案】亲切的怀恋 一切都是瞬息，一切都将会过去；而那过去了的，就会成为亲切的怀恋。 ——题记学习之余，我坐在窗边遐想，想我那遥远的未来、我那充实的现实和那往日的点点滴滴。回忆涌现出的都是那亲切的怀恋。 那是一个阳光明媚的早晨，姥姥教我骑自行车。笨拙的我却驾驭不了这高傲的自行车。我一次次地从自行车上摔落，姥姥一次次将我扶起。她不厌其烦地教我骑自行车的方法，又
E X C E L中的绘图工具 使用技巧 Document serial number【LGGKGB-LGG98YT-LGGT8CB-LGUT-
EXCEL中的绘图工具使用技巧 2009-04-17 11:59:37 阅读5396 评论1 字号：大中小订阅 基础班选修课讲义 1、从绘图工具栏直接绘制直线、单边带箭头直线、椭圆、矩形； 2、从绘图工具栏“自选图形”中选择：线条、基本形状、箭头汇总、流程图、星与旗帜、标注、其它自选图形…… 3、添加和编辑绘制图形中的文字内容（右键－添加文字）； 4、工具栏中直接选择“文本框”、“竖排文本框”； 5、图形右键菜单的其它功能（超链接、宏等）。 二、绘制单位正方形、单位圆、单位正xx…… 1、从绘图工具栏及其“自选图形”中选择矩形、椭圆、菱形、平行四边形等图标……到单元格内点击一下，就可生成单位正方形、单位圆、单位正xx……（对线条不起作用）； 2、在绘图工具栏的“自选图形”中双击选择的图形，可直接生成单位圆、单位正方形、单位正菱形、单位正平行四边形等图案（单位边长）。 三、连续绘图 1、双击绘图工具栏中需要绘制的图形，比如椭圆、长方形、直线，以后可以连续绘制出椭圆/圆、长方形/正方形以及直线；按ESC键取消连续画图； 2、对绘图工具栏“自选图形”中的图形使用双击，只能产生“单位正xx”图形，不能连续绘图（其作用见“二、2”条）。 3、右键绘图工具栏－自定义－命令－自选图形－选择命令－拖到工具栏中，今后就可以双击这些图形连续绘图。 四、改变图形 1、改变绘制图形的线条、箭头、边框、色彩设置（右键或者双击图形，设置自选图形格式；工具栏中选择）； 2、点击和转动相关图形中的绿色圆点转动图形； 3、使用绘图工具栏“翻转”按钮； 4、点击和拖动相关图形中的黄色圆点改变图形； 5、右键已经画好的图形，点击绘图工具栏的“绘图”，选择“改变自选图形”，再选择你想得到的图形； 6、绘图默认效果的设置； 7、（重要应用之一）改变单元格批注图形：单元格批注显示后，右键“批示”框，点击绘图工具栏的“绘图”，选择“改变自选图形”，再选择你要得到的图形；
常用机械制图手工绘图工具及使用 技巧 熟练掌握常用的绘图工具使用技巧,对于提高手工绘图的质量和速率有重要意义。 —、常用绘图工具 （1）（图板）画图时，需将图纸平铺在图纸上，所以，图板的表面必须平整、光洁、且富有弹性。图板 的左侧边称为导边，必须平直。常用的图板规格有0号、1号和二号三种。 （2）丁字尺丁字尺主要用于画水平线，它由尺头和尺身组成。尺头和尺身的连接处必须牢固，尺头的 内侧边与尺身的上边（称为工作边）必须垂直。使用时，用左手扶住尺头，将尺头的内侧边紧贴图板的 导边，上下移动丁字尺，自左向右可画出一系列不同位置的水平线，如图1–18a所示。 （3）三角板三角板有45°-90°角和30°-60°-90°角的各一块。将一块三角板与丁字尺配合使用，自下而上 可画出一系列不同位置的直线，如图1-18b所示；还可画与水平线成特殊角度如30°、45°、60°的倾斜线，如图1-18c所示将两快三角板与丁字尺配合使用，可画出与水平线成15°、75°的倾斜线，如图2所示。两块三角板互相配合使用，可任画已知直线的水平线或垂直线，如图3所示。 图1用丁字尺和三角板画线 图2画15度75度斜线 图3画已知直线平行线和垂直线 二、分规、比例尺 (1)分规分规是用来量取尺寸、截取线段、等分线段的工具。分规的两腿端部有钢针，当两腿合龙时， 两针尖应重合于一点，如图4所示。图5所示为用分规在比例尺上量取尺寸（图5a），然后在线上连续截取等长线段（图5b）的方法若欲将图5c所示的AB线段四等分，可先任凭自测估计，将分规的两针 尖开到约为AB/4进行试分，如有剩余（或不足）时，再将针尖间的距离张大（或缩小）e/4，e为剩余或不足量，再进行试分，直到满意为止。用试分法也可等分圆或圆弧。 (2)比例尺比例尺的三个棱面上有六种不同比例的刻度，如1：100、1：200等，可用于量取不同比例的 尺寸。 图5分规画法 三、圆规圆规是用来画圆或圆弧的工具。圆规固定腿上的钢针具有两种不同形状的尖端：带台阶的尖端是画圆货圆弧时定心用的；带锥形的尖端可作分规使用。活动腿上有肘形关节，可随时装换铅芯插脚、 鸭嘴脚及作分规用的锥形钢针插脚，如图6所示。 图6圆规及附件 画圆或圆弧时，要注意调整钢针在固定腿上的位置，使两腿在合龙时针尖比铅芯稍长些，以便将针尖全部扎入内，如图7a所示；按顺时针方向转动圆规，并稍向前倾斜，此时，要保证针尖和笔尖均垂直纸 面，如图7b所示；画大圆时，可接上延长杆后使用，如图7c所示。 图7圆规用法 四、曲线板曲线板是绘制非圆曲线的常用工具。画线时，先徒手将各点轻轻地连成曲线，如图8a所示；然后在曲线板上选取曲率相当的部分，分几段逐次将各点连成曲线，但每段都不要全部描完，至少留出后两点间的一小段，使之与下段吻合，以保证曲线的光滑连接，如图8b所示。 图8非圆曲线的描绘 五、铅笔(1)铅笔的型号及应用绘图铅笔分软与硬两种型号，字母“B”表示软铅笔，字母“H”表示硬铅芯。“B” 之前的数值越大，表示铅芯越硬。 之前的数值越大，表示铅芯越软；“H” 字母“HB”表示软硬适中的铅芯。 图9修磨铅笔的方法
人不可貌相作文 人不可貌相 最近参加一个活动，认识了一位老板，这位老板是来当地考察投资的。我接过他递上的名片，是那种带花纹的发亮的名片，感觉很俗。上面列了好几家公司，密密麻麻的。我也没细看，总之他是个有钱人。他客气地跟我说请我去他那儿玩儿，可以住他开的宾馆。我敷衍两句就把名片放包里了。老板说一口闽南话，闽南话在很多时候就像是老板的专用方言，因为影视剧里的老板大多说闽南话。于是凭一张名片和一口闽南话，我感觉，我跟这人完全不可能说到一块儿去。他不就是一个会挣钱的老板吗？ 那天早上我们要去山里，天气很冷，我穿了薄毛衣还带了件风衣，而他只穿了短袖T恤。在主人的一再劝说下，他在路边的一家小店买了一件运动外套。上车后他解释说，虽然他是福建人，但不怕冷，因为他在东北当过兵。我暗暗吃了一惊，问他是哪年兵，他说1980年。虽然比我晚几年，也是老兵了。接着他又说，虽然当兵只有两三年，但至今依然保持着早上6点起床的习惯，从不睡懒觉。我笑了，感到一丝亲切。 也许是因为我的笑容和语气，他主动跟我聊起来，他说来这个山区考察，除了生意外，也是想做些善事。我有些意外和不解。他略略有些动情地说：“我成为今天这个样子，是受了两个人的影响——一个是乞丐，一个是老板。” ————来源网络搜集整理，仅供个人学习查参考
我不清楚他说“成为今天这个样子”是个什么样子，他的表达不是那么准确，但我已经有了与他聊天的欲望。 他说，刚离开部队的头几年，还处于创业阶段时，他去云南出差，途中在一家小饭店吃饭。等菜的时候，看到饭店老板在撵一个要饭的，很凶，也许是怕要饭的会影响饭店生意。那乞丐面黄肌瘦，被撵后战战兢兢。他看不过去，就拿出5元钱给老板，说你给他一碗肉吃吧。老板就盛了一大碗肉端到门外给那要饭的，没想到门外还有五六个乞丐，他们狼吞虎咽地分享了那碗肉，然后进门来给他作揖，他们站在他的面前，不停地作揖，嘴里喃喃道“恩人”。他说，那个时候他心酸得没法说，忍着眼泪摆手让他们走。 “就是一碗肉啊，他们差不多要把我当菩萨了。这件事改变了我的人生。第一，我想我要努力挣钱，不能过苦日子；第二，我想我挣了钱以后，一定要帮助穷人。” 我可以想象那样的场景。虽然我们都知道，今天仍有很多穷苦的人与我们同处一片天空下，但这个“知道”是抽象的，当他们非常具象地出现在面前时，那种震撼是完全不同的。 老板接着说：“第二个影响我的人是一个台湾老板。这些年我生意慢慢做大了，条件好了，差不多要忘了那个乞丐。因为做生意，我跟一个台湾老板有交往，他回到福建老家做了很多善事，捐建学校，捐修公路，资助穷苦学生，大笔大笔的钱拿出来。可是我发现他跟他的老伴非常节俭，每次过来谈生意，都是自己带着馒头和咸菜，连矿泉水都舍不得买，自己带水壶；穿着也很朴素，从来不穿名牌，而且总是住
八年级作文范文汇编10 篇 一、作文写作 1．根据要求作文。 目： ______如此美 要求：① 在横上添上恰当的，将目充完整；②除歌外，文体不限。③有真 情感，不得套写抄；④用行范的言文字表达；⑤不少于600 字；⑥文章中不 要出真的地名、校名和人名。 【答案】参考例文 微笑如此美 微笑是最好的止痛。人生在世，失在所免，正因有成功的喜悦与失的苦楚，才成了我五 彩的生活。 —— 步入初中以来，人的最大感就是考多，平凡的考，不免有所失意。一眨眼，月考来了，尽 管之前有很多考，但却只是在班上。面一次全校大考，提心吊胆是不必的，然而我却在如 此重要的考中失意了。 那是一个阴森的下午，老在教室里着考的排名，久久未能听我的名字，一个个熟悉的人 名，活像一把把陌生的尖刀，刺痛着我的心。那一卷就像一沉重的石，将我入无底的深 渊。下了，教室里的人影越来越少，最后独留我一个。 一人在与世隔之地，我迷茫的望着窗外，一云遮住了天，太阳同也遮住了我的心。 不知什么候，身后响起了脚步声，越来越清晰，似乎正在向我靠。我不由自主的扭去，是 他，他上着从未停歇的笑容是那么的，以至于我忘了前一秒的 痛。慢慢的，他走到我的身前，并未我感到惊，因他是我最知心的朋友阿A。 很久很久，我都没，只是他始微笑着我。我也不知所措的看着他。目光交的 一刹那，我看到了一种从未有的温暖，加上他那始不的笑容，更像是一熊熊烈火点燃了 我心底最后一希望。 很久之后，他于了：“不要苦着，笑一笑走你的愁吧。”尽管只是一句毫不起眼的，瞬我 体会到了笑的魔力，我的上也开始洋溢微笑，的。 云始未能抵住阳光的威力，我心中的那一片云也消散在微笑之中。微笑是如此的神奇， 又是如此的美，以至于医治好了我心中的痛。 【解析】 【解】 是一道半命作文，作文既了写作情境，又限定了写作范；比全命作文灵活，比作文 更具束性，具有收放自如的特点。半命作文比突出地考了学生的、构思及言表达等多 方面的能力。写作注意以下几点： ①要意，才能确定好的立意。理解“美”的含，“美”，可以从不同角度理解： (1) 具体事物 之美。可以从自己熟悉的人、事、物、景中取一种，而且必是“珍品”来写人、事、物、景的美 或外表、外形的美，如校园如此美，落叶如此美， 某人如此美，家如此美，春天如此美，大海如此美，??(2) 精神面的美：
Puritans and Harvard 清教徒和哈佛 Puritans played a very important role in Harvard history as well as in American history. 清教徒不仅在美国历史中扮演过重要角色，同样在哈佛大学的历史中也起到了重要作用。 In the 16th and 17th century, a series of religious reforms took place in England Finally, the Church of England was established as the Established Church. People did not have religious freedom. Those people who did not agree with the Church of England were regarded as Separatists and were persecuted. To escape religious persecution, many people fled from England to other countries. Many Puritans chose to go to the New World. 在16和17世纪的英国发生了一系列宗教改革，最终圣公会被定为英国国教。人们没有宗教自由。那些不同意英国国教观点的人被视为宗教分裂者，并遭到迫害。为了逃脱宗教迫害，许多人逃离英国，前往其他国家。很多清教徒选择移民到新大陆。 In 1620, a ship named Mayflower left England. It transported the English Separatists, better known as Pilgrims, from Plymouth, England to Plymouth, Massachusetts, the United States. There were 102 passengers, more than one third of whom were Puritans. During the following decade, the number of Puritans grew. Then in 1628 these Puritans established the Massachusetts Bay Colony. 1620年，一艘名为“五月花号”的船离开了英国，它载着英国宗教分离派——更熟悉的称号是“清教徒前辈移民”，离开了英国的普利茅斯前往美国马萨诸塞州的普利茅斯。船上有102名乘客，其中1/3多的人是清教徒。在接下来的10年中，清教徒队伍逐渐壮大起来，并于1628年建立了马萨诸塞湾殖民地。 Many Puritans had received classic style of higher education in Oxford University and Cambridge University in England. They wanted to pass on to their descendants this kind of education. Besides, these colonists saw colleges as an effective way to disperse religious belief. Therefore, in 1636 the first and oldest institution of higher learning in the United States was established by vote of the Great and General Court of Massachusetts Bay Colony, 140 years earlier than the foundation of the United States. 许多清教徒都曾在英国的牛津和剑桥大学接受过古典式的高等教育，他们想把这种教育传给子孙后代。此外，这些殖民者把大学看做是传播宗教信仰的有效途径。所以，1636年马萨诸塞湾殖民地法院投票通过并建立了美国第一所也是最古老的一所高等教育机构，比美国建国早了140年。 The institution was initially named “New College” or “the College at New Towne"（or “Cambridge College”). The town where the college located was named Cambridge, becau se many Puritans had studied in Cambridge University. The name of the town demonstrated that many people still cherished their memory of life in England. Although they had been persecuted in England, many people still saw England as their motherland. The college followed the style of English universities as well. Then in 1639 the college was renamed Harvard College after a benefactor called John Harvard who was a clergyman. Therefore, in a sense, Harvard University was the product of religious activities.