Local image display

Open or close the local image display window. An PatInspect tool supports the following image buffers:

• The Current.InputImage buffer contains the current image and the region defining the portion of the image you want the tool to analyze.
• The Current.TrainImage buffer contains the training image you get from the PMAlign tool.
• The Current.TrainedPatternImage buffer contains the trained pattern, created either from a single training image or multiple training images.
• The Current.TrainedThresholdImage buffer contains the current threshold image.
• The LastRun.InputImage buffer contains the last image processed by the PatInspect tool along with any graphics you enable from the Graphics tab.
• The LastRun.MatchImage buffer contains the portion of the run-time image the PatInspect tool compared against the trained pattern. The tool can generate a normalized and non-normalized match image.
• The LastRun.DifferenceImageAbsolute buffer contains the thresholded difference image. The tool can also generate a thresholded difference image containing only the brighter pixels or a thresholded difference image containing only the darker pixels.
• The LastRun.DifferenceImageRawAbsolute buffer contains the raw difference image. The tool can also generate a raw difference image containing only the brighter pixels or a raw difference image containing only the darker pixels.

Reset

Reset the vision tool to its default state.

This tool gives you a choice between resetting to the default-constructed state, which is appropriate when you are using it in a Visual Studio.NET application, and its template-initialized state, which is appropriate for QuickBuild applications.

SobelScale and SobelOffset

The tool uses these Sobel coefficients to generate a pseudo standard deviation image when you create the trained pattern from a single training image. If a PatInspect tool consistently reports false defects with outlines or borders of your run-time images, try adjusting the Sobel coefficient values.

Changing the Sobel coefficient values has no effect if you use the statistical training method to create the trained pattern.

Grab Train Image and Origin

Copy the current input image, and the origin as reported by the PMAlign tool, into the Current.TrainImage buffer. Once you have a training image you can switch to the Current.TrainImage buffer and set the region of interest within the image that you want the tool to examine.

This button is not available until you have an input image, and becomes disabled once you have created a trained pattern.

Train

Create a trained pattern from the portion of the training image currently enclosed within the region of interest.

This button is not available until you have a training image in the Current.TrainImage buffer, and becomes disabled once you have created a trained pattern from a single image. To create a trained pattern using statistical training, you need to acquire another input image and use the Statistically Train Current Pattern button.

StatisticalTrain

Incorporate the current input image into the existing trained pattern.

You should acquire a new input image and obtain a new Pose before pressing this button each time. Pressing this button repeatedly for the same image and the same Pose typically results in a poorly-trained pattern. See the topic How To Use a PatInspect Tool for an example of capturing multiple images to train a pattern.

EndStatisticalTraining

End statistical training on this pattern.

Once you click this button, no further statistical training can be performed. Ending stastical training will free some internal data, causing this vision tool to use less memory.

TrainRegionMode

Define how the tool will interpret the training region you specify. Choose any of the following options:

• Pixel Aligned Bounding Box processes all of the input image within a bounding box surrounding the region shape you are currently using.
• Pixel Aligned Bounding Box Adjust Mask encloses the defined region within a rectangle, but masks out the image portions that are outside the region but inside the enclosing rectangle. The result is that the defined region is closer to what you specify.
• Affine Transform is valid only if you specify a CogRectangleAffine for the region shape, where the pixels within the affine rectangle undergo an affine transformation to remove the effect of any rotation and skew before the tool attempts to decode the symbol.

TrainRegion

Select the shape of the input region. Selecting "None=Use entire image" means that the tool uses the entire input image. PatInspect tool supports the following input region shapes:

• CogCircle
• CogEllipse
• CogPolygon
• CogRectangle
• CogRectangleAffine

The set of region-defining parameters that appear depend on the region shape you use. For more information on using a polygon as an input region, see the topic Using Polygon Input Regions.

TailHigh

Set the percentage of pixels to discard on the high end of the grey scale spectrum before performing the match tails method of image normalization.

Note: Only shown if MatchTails is selected for the mode.

TailLow

Set the percentage of pixels to discard on the low end of the grey scale spectrum before performing the match tails method of image normalization.

Note: Only shown if MatchTails is selected for the mode.

NeighborhoodWidth

The width of the neighborhood, in pixels, to use for the local correction normalization modes.

Note: Only shown if LocalCorrection or LocalCorrectionEnhanced is selected for the mode.

NeighborhoodHeight

The height of the neighborhood, in pixels, to use for the local correction normalization modes.

Note: Only shown if LocalCorrection or LocalCorrectionEnhanced is selected for the mode.