The tab Usage allows you to apply your model on test images and adapt suitable search parameters.
Determine the source type of your test images.
| Menu Entry | Description |
| Image Files | Test images are loaded from file. |
| Image Acquisition Assistant | Use the 图像采集助手 to acquire your test images. |
You can select a test image from the text box of the dialog Test Images. The selected image is automatically displayed in the graphics window of HDevelop.
| Menu Entry | Description |
| Load/Add | Load an image from file or add an image using the 图像采集助手. |
| Live Image | Add the image acquired by the live image acquisition mode of the 图像采集助手. |
| Remove | Remove the currently selected test image from the list. |
| Remove All | Remove all test images. |
| Save | Save the selected image if it has been added using the 图像采集助手. |
| Save All | Save all images added by the 图像采集助手. |
| Set Reference | Set the match in the chosen image as reference position for
alignment. Note: This is only possible if a match is detected in the image. If no reference image is chosen, the model image will be used as basis for the alignment or rectfication. |
| Detect All | Search all objects in the complete sequence of test images that were
loaded before (independent of the set
Maximum Number of Matches).
The results are displayed successively in the graphics window. Note: For descriptor-based matching this also sets the Maximum Number of Matches if it has not been set previously (it is set to 0). Note: This also sets the number of visible objects in the test image if it has not been set previously. Note: If the button is clicked for the first time or after you changed a model parameter, the internally stored model is actually created, which takes some time. |
| Find Model | Search the object in the currently
selected test image.
The result is displayed in the assistant window. Note: If the button is clicked for the first time or after you changed a model parameter, the internally stored model is actually created, which takes some time. |
| Always Find | Start the matching process automatically on the selected test image.
The result is displayed in the graphics window. Note: If the matching process is started for the first time of after you changed a model parameter, the internally stored model is actually created, which takes some time. |
| Number of Visible Objects in Test Image | Specify how many objects are really visible in the current test image. To do so, click onto the currently displayed number of detected objects (Visible) in the text field of the currently selected test image. The specified numbers of visible objects are used when determining the recognition rate, that is, the recognition rate is 100% when the sum of all objects found in the test images is equal to the sum of the specified numbers. |
The menu item Standard Use Parameters allows you to specify search parameters.
The following table gives you an overview over the available standard search parameters for each matching method: The abbreviations used are SBM for shape-based matching, CBM for correlation-based matching, DBM for descriptor-based matching, and DM for deformable matching.
| Menu Entry | Description |
| Starting Angle | SBM: ❌, CBM: ✓, DBM: ✓, DM: ✓ Specify the starting angle of the allowed range of rotation (unit: °). Example: To allow model rotations up to +/-5 °, for example, you should set the starting angle to -5 ° and the angle extent to 10 ° or angle end to 5 °. Note: The range of rotation is defined relative to the created model. For a model created from an image a starting angle of 0 ° corresponds to the orientation the object has in the model image. |
| Ending Angle | SBM: ❌, CBM: ✓, DBM: ✓, DM: ✓ Specify the ending angle of the allowed range of rotation (unit: °). Example: To allow model rotations up to +/-5 °, for example, you should set the ending angle to 5 ° and the starting angle extent to -5 °. Note: The range of rotation is defined relative to the created model. For a model created from an image an ending angle of 0 ° corresponds to the orientation the object has in the model image. |
| Minimum Score | SBM: ❌, CBM: ❌, DBM: ❌, DM: ❌ Specify the minimum score that a potential match must reach to be regarded as model instance in the image. For further information see Solution Guide II-B: Matching. |
| Maximum Number of Matches | SBM: ❌, CBM: ❌, DBM: ❌, DM: ❌ Specify how many instances of the object are searched for in the image. If you specify the value 0, all found instances are displayed. Note: The parameter sets a maximum value, that is, if more object instances are present in the image only the best instances of the specified number are displayed. |
The menu item Advanced Use Parameters allows you to specify search parameters.
The following table gives you an overview over the available advanced search parameters for each matching method: The abbreviations used are SBM for shape-based matching, CBM for correlation-based matching, DBM for descriptor-based matching, and DM for deformable matching.
| Menu Entry | Description |
| Greediness | SBM: ❌, CBM: ✓, DBM: ✓, DM: ❌ Determine how 'greedily' the search should be carried out. For more information see Solution Guide II-B: Matching. |
| Maximum Overlap | SBM: ❌, CBM: ✓, DBM: ✓, DM: ❌ Determines by what fraction two instances may overlap at most in order to be considered as different instances and hence to be returned separately. For a schema see Solution Guide II-B: Matching. |
| Subpixel | SBM: ❌, CBM: ❌, DBM: ✓, DM: ❌ Determine which type of subpixel refinement should be carried out. For more information and possible values see Solution Guide II-B: Matching, (shape-based matching) find_ncc_model (correlation-based matching), and find_planar_uncalib_deformable_model (deformable matching). |
| Max Deformation | SBM: ❌, CBM: ✓, DBM: ✓, DM: ✓ Determine by how much an object is allowed to deviate from the model in order to be considered as a match. The maximal allowable object deformation is specified in pixels. For more information see Solution Guide II-B: Matching. |
| Last Pyramid Level | SBM: ❌, CBM: ❌, DBM: ✓, DM: ❌ Determine the lowest pyramid level to which the found matches are tracked. For more information on the image pyramid see Solution Guide II-B: Matching. Note: The lowest pyramid level is denoted by a value of 1 (shape-based and deformable) and 0 (correlation-based), respectively. Example: When selecting the value 2, the matching starts at the highest pyramid level and tracks the matches to the second lowest pyramid level. |
| Timeout | SBM: ❌, CBM: ❌, DBM: ✓, DM: ✓ Set a maximum runtime of the operators used to find the shape model. To use the Timeout function, activate Enable on the right side of the tab. Then choose the time in milliseconds after which the search for a model is aborted. |
| Shape models may cross the image border | SBM: ❌, CBM: ✓, DBM: ✓, DM: ✓ Determine whether the models to be found may lie partially outside the image, that is, whether they may cross the image border, independent of the domain. For more information see the description of the parameter border_shape_models of set_generic_shape_model_param. |
| Robust Pyramid Tracking | SBM: ❌, CBM: ✓, DBM: ✓, DM: ✓ Enable a mode to automatically detect during the search the lowest image pyramid level on which at least one match can be found. For more information see the description of the parameter pyramid_level_robust_tracking of set_generic_shape_model_param. |
| Score Type | SBM: ✓, CBM: ✓, DBM: ❌, DM: ✓ Define the type of score to be evaluated. For more information on the parameter and possible values see Solution Guide II-B: Matching. |
| Descriptor Min. Score | SBM: ✓, CBM: ✓, DBM: ❌, DM: ✓ Determine what score a potential match must at least have to be regarded as an instance of the model in the image. For more information on the parameter see Solution Guide II-B: Matching. |
| Guided Matching | SBM: ✓, CBM: ✓, DBM: ❌, DM: ✓ Enable the usage of an already approximately known homography to 'guide' the matching process, which enhances the accuracy of the object recognition. For more information on the parameter see Solution Guide II-B: Matching. |
This section allows you to optimizes the recognition speed by means of automatically determining the values for the parameters Minimum Score and Greediness.
How it works: At the beginning, the Greediness is set to 0 and the Minimum Score to 1. Then, the minimum score is decreased until the matching succeeds in all test images, that is, until the recognition rate is 100%. Now, the Greediness is increased as long as the matching succeeds. This process is repeated until the optimum parameters are found. You can lower the threshold of acceptance for the recognition rate manually using the corresponding slider or text box at the bottom of the dialog. The Matching Assistant then displays the optimal values for Minimum Score and Greediness and the reached recognition time. The found values are set as new search parameters.
The speed is calculated as the average recognition speed over all test images.
| Menu Entry | Description |
| Run Optimization | Run the optimization. |
| Stop | Interrupt the process of optimizing the recognition speed; please note however, that this event is processed only after the current search has finished. |
| Mode | Specify how the recognition rate is defined.
The following options are supported:
|
| Recognition Rate | Determine the required recognition rate. |