The following table summarizes the parameters that you supply to PatMax and the results that PatMax returns to you. Different parameters and results are used and returned for training and alignment.
| Type | Item | Notes |
|---|---|---|
| Training parameters | Model origin |
Location in pattern model to use when returning result locations in run-time images
Can be set at run time with minimal time penalty. |
| Training image | The pattern of interest | |
| Training mode | Whether to train using image pixels or shape models | |
| Elasticity |
Amount of tolerance (in pixels) for nonlinear geometric change in feature location
Can be set at run time with minimal time penalty. |
|
| Polarity tolerance |
Either consider or ignore the polarity information in patterns in the run-time image
Can be set at run time with minimal time penalty. |
|
| Feature granularity |
Controls the size of features that make up a trained pattern. Must set before training.
Do not set manually for most applications. |
|
| High sensitivity mode | Use standard mode for normal images, high sensitivity mode for noisy or low contrast images. | |
| Sensitivity parameter | Use the default value for most applications | |
| Mask image | Excludes parts of training image from pattern | |
| Repeating patterns | Allows for repeating elements in the pattern you want to train | |
| Auto edge threshold | Disables the automatic edge threshold in favor of another value your images require | |
| Alignment parameters | Enabled degrees of freedom | Only enabled transformations are measured |
| Nominal transformation value(s) and zone(s) | For each non-enabled degree of freedom, specify a nominal value. For each enabled degree of freedom, specify a range of values (zone). | |
| Number of results | The number of instances of the pattern you expect PatMax to find in the run-time image | |
| Contrast threshold | The minimum contrast value for pattern instances | |
| Score using clutter | Whether or not to consider the effect of extraneous features on pattern instance score | |
| Accept threshold | Threshold for considering an instance of the pattern as valid | |
| Mask image | Exclude parts of run-time image from consideration | |
| Timeout | Maximum amount of run time PatMax use | |
| Coarse Accept threshold | Change the default value the tool uses to determine the coarse features of your pattern in run-time images, which will vary the number of potential candidates the tool will use to refine the search. | |
| Alignment results | Location | Location of each instance |
| Pose | Description of how instance is mapped from trained pattern to run-time pattern instance | |
| Scores | Indication of degree of fit |