当前位置:文档之家› eCognition_Orientation_Training

eCognition_Orientation_Training

Definiens Developer, Definiens eCognition? Server
Orientation Training
Workflow Overview, main features and functions

Imprint and Version
Document Version Copyright ? 2008 Definiens AG. All rights reserved. Published by Definiens AG Trappentreustr. 1 D-80339 München Germany Phone Fax Web +49-89-231180-0 +49-89-231180-90 https://www.doczj.com/doc/a014076726.html,
Legal Notes
Definiens?, Definiens Cellenger? and Definiens Cognition Network Technology? are registered trademarks of Definiens AG in Germany and other countries. Cognition Network Technology?, Definiens eCognition?, Enterprise Image Intelligence?, and Understanding Images?, are trademarks of Definiens AG in Germany and other countries. All other product names, company names, and brand names mentioned in this document may be trademark properties of their respective holders. Protected by patents US 7146380, US 7117131, US 6832002, US 6738513, US 6229920, US 6091852, EP 0863485, WO 00/54176, WO 00/60497, WO 00/63788 WO 01/45033, WO 01/71577, WO 01/75574, and WO 02/05198. Further patents pending.

Table of Contents
Orientation Training______________________________________________________ 1 Imprint and Version __________________________________________________ 2 Legal Notes _________________________________________________________ 2 Table of Contents ________________________________________________________ 3 Module 1 Introduction to the Definiens Product Family ______________ 5 Definiens Enterprise Image Intelligence Suite - Client Software Feature Comparison 6 Module 2 Goal of the Workshop __________________________________ 9 Lesson 2.1 Introductions __________________________________________ 9 2.1.1 Key analysis 9 2.1.2 Workspace Automation Framework 10 2.1.3 Input Data 11 2.1.4 Landcover Classification 11 2.1.5 Shrub Density Classification 11 Lesson 2.2 View Input Data _______________________________________ 12 2.2.1 Open the Workspace 12 2.2.2 Open the unprocessed Project File 13 2.2.3 Changing the Layer Mixing 13 Lesson 2.3 Start the Batch Analysis_________________________________ 15 2.3.1 Open the ‘Start Analysis’ dialog box 15 2.3.2 Define the ‘Job Scheduler’ location 16 2.3.3 Define the Rule Set 16 2.3.4 Check the Settings for Export 16 2.3.5 Execute the Analysis 16 Lesson 2.4 View the Result of the Analysis ___________________________ 18 2.4.1 Open the Processed Subset 18 2.4.2 Visualize Image Object 18 2.4.3 Display Classification results 20 2.4.4 Panning and Zooming Functions 21 Module 3 Reproduce the Individual Classification Steps ____________ 23 Lesson 3.1 Recreate the State before Landcover and Shrub Density Analysis 23 Lesson 3.2 Prepare Settings and Load the Rule Set ____________________ 24 3.2.1 Change to the predefined view for Rule Set Development 24 3.2.2 Load the Rule Set and examine the subroutines 25 Lesson 3.3 Overview over Analysis steps in Subroutine ‘OnStitched ______ 27 3.3.1 Remove the Tile Border Effects: 27 3.3.2 Landcover Classification 28 3.3.3 Heat Map 28 3.3.4 Export 29 Lesson 3.4 Execute the Border Effect Elimination _____________________ 29 Lesson 3.5 Examine the Landcover Classification Processes _____________ 30 3.5.1 Water Body 30 3.5.2 Vegetation 32 3.5.3 Bare Soil and Impervious 33 3.5.4 Shrub 37 Lesson 3.6 Classify Density Steps __________________________________ 38 3.6.1 Grow Objects for Neighborhood Analsyis 38 3.6.2 Get Number of Shrub Neighbors 40 3.6.3 Classify Shrub Density Classes 40 3.6.4 Generalize Shrub Density Objects 41 Module 4 Hands on Rule Set Development on small subset__________ 45


Definiens Developer, Definiens eCognition? Server Orientation Training
Introduction to the Definiens Product Family
Module 1 Introduction to the Definiens Product Family
The Definiens Enterprise Image Intelligence? Suite is a comprehensive image analysis platform. It contains all the client and server software needed to extract intelligence from any digital image in a fully-automated or semi-automated way. The client software is role-based and supports the needs and skills of different users in an organization. The server software, known as the Definiens eCognition? Server, is a processing environment that allows the batch processing of jobs. It is hugely scalable, capable of handling millions of images in a single job.
5

Definiens Developer, Definiens eCognition? Server Orientation Training
Introduction to the Definiens Product Family
Definiens Enterprise Image Intelligence Suite? - Client Software Feature Comparison
The Definiens Enterprise Image Intelligence? Suite is the platform on which image analysis applications are built and executed. It comprises role-based products which meet the needs of any image-based business. The table below illustrates a comparison between the different clients and their features: Definiens Definiens Developer Architect Develop & test analysis ruleware Create & package solutions Create custom importers Import data Analyze images locally Submit batch analysis jobs Calibrate solutions for image conditions Solution workflow builder Semi-automated workflow Fully automated workflow Review results Manage grid hardware Install & deploy software Manage data stores Manage users & security Web-based X X X X X X X X X X Definiens Analyst Definiens Viewer Definiens Administration Console
X
X
X X X X
X X X X X X X X X X X X
6

Definiens Developer, Definiens eCognition? Server Orientation Training Definiens Developer
Introduction to the Definiens Product Family
Definiens Developer is a powerful and completely integrated environment designed for image analysis specialists to develop, test and package new image analysis applications. Definiens Developer can be used as a standalone tool or in combination with the Definiens eCognition? Server. Definiens Architect Definiens Architect is a simple end user tool which enables domain experts and other users to calibrate and run pre-packaged image analysis applications. Definiens Analyst Definiens Analyst enables image analysts and other users to import data, run image analysis applications within the Definiens eCognition? Server and review the results. Definiens Viewer Definiens Viewer enables managers and other users to review the results of every image analysis conducted via the Definiens Enterprise Image Intelligence? Suite. Definiens eCognition? Server The Definiens eCognition? Server provides a processing environment for the batch execution of image analysis jobs using a high-performance grid computing environment. Definiens Administration Console The Definiens Administration Console provides administrators with a web-based interface that simplifies the management of the Definiens eCognition? Server environment. Definiens Data Management Definiens Data Management offers an open, enterprise ready and cost effective solution for managing the huge volume of data generated by image analysis projects. The data is managed using standard relational database technologies. Definiens Software Development Kit (SDK) The Definiens Software Development Kit (SDK) enables the integration of the Definiens Enterprise Image Intelligence? Suite within any business process, using any data source or target, and allows the core analysis capabilities of the Definiens eCognition? Server to be extended.
7

Definiens Developer, Definiens eCognition? Server Orientation Training
Introduction to the Definiens Product Family
8

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
Module 2 Goal of the Workshop
This Module has the following Lessons:
Error! Reference source not found. View Input Data Start the Batch Analysis View the Result of the Analysis Goal of this Workshop is to present a shrub density analysis on large Quickbird data (panchromatic and multispectral) using Definiens Image Analysis Products.
Lesson 2.1 Introductions
This Lesson has the following content:
Introduction to the Key Analysis Introduction to the Workspace Automation Framework Input Data Landcover Classification Shrub Density Classification
2.1.1 Introduction to the Key Analysis
The actual image classification is done by analyzing in a first step the landcover, including the shrubs, and in a second step the density of those shrubs. The key feature to do this is using the number of classified shrubs in a certain neighborhood as criteria. 5 Density steps are evaluated.
9

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
2.1.2 Introduction to the Workspace Automation Framework
To provide maximum processing power the Definiens Workspace Automation concept is used and therefore Definiens Developer and Definiens eCognition? Server is used. This means, the Rule Set contains a combination of actual segmentation and classification commands and commands to execute the tiling of the image and the stitching of the results after processing. Overall workflow:
N Tiles are created; OTiles are analyzed by running a segmentation algorithm to create image objects; PTiles are stitched back to the subset again together their with image objects; QThe stitched subset is processed with the Landcover and Shrub Density Analysis
N P Q
O
10

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
2.1.3 Input Data
2.1.4 Landcover Classification
2.1.5 Shrub Density Classification
11

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
Lesson 2.2 View Input Data
This Lesson has the following content:
Open the Workspace Open the unprocessed Project File Changing the Layer Mixing
2.2.1 Open the Workspace
Methodology A Workspace file contains image data references, projects, exported result parameters and references to the used rule set. Furthermore, it comprises the import and export templates, result states, and metadata. In the Workspace window, you administer the workspace files. Here you manage all relevant data of your image analysis tasks. 1. Action! From the main menu select ‘File>Open Workspace’ or press the ‘Open Workspace’ button 2. .
Browse to the ‘01_DEV_7_ORIENTATION\WS_Shrub_Density_Analysis’ folder where the training data is stored and select the ‘WS_Shrub_Density_Analysis.dpj’ file. Select the topmost item in the left Workspace window ‘WS_Shrub_Density_Analysis’.
3.
The Workspace window is split in two panes: Methodology ? The left pane contains the Workspace tree view. It is possible to structure the data also in several different folders. The hierarchical structure of the folders that contain the projects is represented then in this window. Click a plus sign (+) to expand a folder. Click a minus sign (–) to collapse a folder. Select any folder by clicking. On the right pane, you see the content of the selected folder. You can choose between the display options Tree View and Folder View or apply filters.
?
For both views, the lower right pane of the Workspace window contains information about the selected project: its state, scale, the time of the last processing and corresponding remarks (if available). Depending on the type of data and the actions run on the data, there are additional columns with information.
Workspace Check
N O
12

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
Content of the loaded Workspace The Workspace contains two projects: ?
N 01_Subset_QB Maricopa with state ‘created’. This means this project is not yet processed. 5856x3777 (0.6 meter resolution; around 13 square kilometers or 5 square miles O 02_QBMaricopa together with its tiles. This project has already been processed,
see column ‘State’ is set to ‘Processed’. The scene has a size of 16996 x 8804 pixels (0.6 meter resolution; around 53 square kilometers or 20 square miles.
?
2.2.2 Open the unprocessed Project File
The project file represents a subset of a whole Quickbird scene (Image data courtesy of DigitalGlobe). This project contains the multispectral channels blue, green red and nir as well as a panchromatic channel. 1. To open a project file from the Workspace window do one of the following: ? ? In the Workspace window double-click the 01_Subset_QB Maricopa project file Right-click on the project in the Workspace window and select Open from the menu. Action! Methodology
NOTE:
The currently opened project is marked with an asterisk in the Workspace window
2.2.3
1. ? ?
Changing the Layer Mixing
From the View menu, select Image Layer Mixing or click on the Edit Image Layer Mixing button toolbar. in the View Settings
To open the Edit Image Layer Mixing, do one of the following:
Figure 1: Layer mixing buttons in the View Settings toolbar
The Edit Image Layer Mixing window opens. The first three bands of data are displayed by default in the Red, Green and Blue color guns.
13

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
Figure 2: Edit Layer Mixing dialog box
2.
To view the image in true color, set the blue, green and red layer to the respective color slots by clicking in the according field.
Figure 3: True color layer mix
3.
To view additionally the nir layer, select the respective box.
Figure 4: False color mix with additional nir layer displayed green
14

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
Result Check
Figure 5: Image data with false color mixing
Lesson 2.3 Start the Batch Analysis
This Lesson has the following content:
Open the ‘Start Analysis’ dialog box Define the ‘Job Scheduler’ location Define the Rule SetCheck the Settings for Export Execute the Analysis Methodology If you have installed the Definiens eCognition? Server software locally or anywhere accessible in the network, you have the possibility, to process your data in batch mode and to use the Workspace Automation Algorithms. This means the actual analysis is not done locally in the Definiens Developer, it is sent to the Definiens eCognition? Server. If there are more servers on more processors installed, parallel processing is possible, the processing time goes down significantly.
2.3.1 Open the ‘Start Analysis’ dialog box
1. 2. Select the 01_Subset_QB Maricopa and right-click on it. Choose ‘Analyze…’ from the list or click the ‘Analyze Projects’ button ‘Tools’ toolbar. in the Action!
The ‘Start Analysis Job’ dialog box opens.
15

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
2.3.2 Define the ‘Job Scheduler’ location
Methodology The ‘Job Scheduler’ is part of the Definiens eCognition? Server installation. The ‘Job Scheduler’ defines the machine as host for a Definiens eCognition? Server. The Job Scheduler performs job queuing and handling. 1. Action! Make sure that the location of the Job Scheduler is defined correctly. In the example above, the Job Scheduler is located locally on the computer.
2.3.3 Define the Rule Set
2. In the ‘Rule Set’ field browse to the ‘01_DEV_7_ORIENTATION’ folder and select the Rule Set file QB_Maricopa_Shrubs.dcp. This is the Rule Set the scenes will be analyzed with.
2.3.4 Check the Settings for Export
Methodology In the ‘Exported Results’ field you can see where which data will be exported. Here in this case the project file, two shape files (HeatMap and Landcover) and a subworkspace will be exported.
Settings Check
Figure 6: ‘Start Analysis Job’ dialog box with Rule Set ‘QB_Shrubs’ defined for analysis.
2.3.5 Execute the Analysis
1. 2. Action! Click on the ‘Start’ button. Check the progress of processing in the Workspace window.
16

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
In the Workspace window you can see the progress of the analysis by checking the State column.
Workspace Check
Figure 7: The State column shows the current status of processing: waiting, processing or processed.
17

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
Lesson 2.4 View the Result of the Analysis
This Lesson has the following content:
Open the Processed Subset Visualize Image Object Display Classification results Panning and Zooming Functions
2.4.1 Open the Processed Subset
1. Action! Double-click on the project as soon as all tiles are processed and the processing of the stitched subset is finished. This indicates that the image data is displayed, not the classification result.
2.4.2 Visualize Image Object
Methodology Image Objects have been created during processing. Image Objects are the building blocks for any further image analysis and also the final result of any analysis. In this project two Image Object Levels do exist. ? ? Level1 representing the Landcover Classification Level 2 representing the Shrub Density Classification
There are several ways to display Image Objects, their contained information or their classification; here the display of their outlines is shown.
Figure 8 View Settings toolbar.
1. Action!
Make sure that the ‘Level1’ is active by selecting it from the drop down menu in the ‘View Navigate’ toolbar.
Figure 9: Level1 set active in the ‘View Navigate’ toolbar.
2.
Make sure that the ‘View Layer’ mode is selected by pressing the
button.
This indicates that the image data is displayed, not the classification result. 3. Select the ‘Show or Hide Outlines’ button to show all outlines of all Image Objects of the level. See how they fit with the image content.
18

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
4. 5.
To show the object mean view select the ‘Pixel View or Object Mean View’ button. To display the outlines of ‘Level1’ select it from the drop down menu in the ‘View Navigate’ toolbar.
Result Check
Figure 10: Image View, Object outlines of Level 1 and Level 2.
19

Definiens Developer, Definiens eCognition? Server Orientation Training
Goal of the Workshop
2.4.3
Methodology
Display Classification results
The Image Objects have been already classified after the conditions defined in the Rule Set. 1. In the ‘View Settings’ toolbar select the ‘View Classification’ button sure that the ‘Show or Hide Outlines’ button is deselected. and make
Action!
2.
Move the cursor over the classification and the assigned class will appear next to it. Select the ‘Pixel View or Objects Mean View’ button transparency view. Select the ‘Show or Hide Outlines’ button the classification colors. to switch on and off the
3.
4.
and the outlines will be displayed in
Result Check
Figure 11: Classification of Landcover in Level 1 and classification of the Shrub Density steps in Level2.
20

相关主题
文本预览
相关文档 最新文档