MimLocatePeak1d

This function detects the position and the intensity of a peak in every lane (row or column) of an image. It is well suited for applications that generate 3D data based on laser line or sheet of light profiling.

The function iterates through each lane. For each lane, the function determines the pixels that are in the peak neighborhood, as defined by PeakWidthNominal, PeakWidthDelta, and MinContrast. For each such region, it finds the position of the highest intensity with sub-pixel accuracy, and calculates the average intensity around that position. The position can be retrieved in a fixed-point representation, with 0 to 7 fractional bits.

For optimal performance, preprocess the result buffer by calling MimLocatePeak1d() with M_PREPROCESS. You can choose to preprocess with or without an image. When preprocessing with an image, specify a typical image in the series of images that you will process with MimLocatePeak1d(). The preprocessed image's size and bit depth are stored. When preprocessing without an image (SrcImageBufId set to M_NULL), you must specify the number of scan lanes in a typical image in the series with ControlFlag. When preprocessing without an image, it is assumed that all images are 8-bit unsigned.

If the preprocess operation is not done explicitly (using M_PREPROCESS), MimLocatePeak1d() will preprocess the result buffer internally when it is first called. If a new image in the series has a different size or bit depth than the image used to preprocess, MimLocatePeak1d() will preprocess the result buffer again, given the new image's characteristics.

In a noisy image, or any image that might have multiple peaks in a single lane, such as when the object or surface is highly reflective, MimLocatePeak1d() can record the position and intensity of multiple peaks in a lane. You can create a custom application that sifts through the multiple peaks in each lane to determine the best peak in the lane for your needs. To record multiple peaks in a lane, use MimControl() with M_NUMBER_OF_PEAKS set to the maximum number of peaks to record. For more information, see the Multiple peaks in a single lane subsection of the Peak intensity detection and depth maps section of Chapter 5: Specialized image processing.

You can retrieve the calculated results from the result buffer using MimGetResultSingle().

MimLocatePeak1d() can accumulate results from multiple frames in one result buffer. To do so, you must set MimControl() with M_NUMBER_OF_FRAMES, which is the maximum number of frames for which results can be accumulated in the result buffer. When calling MimLocatePeak1d(), set M_FRAME_INDEX to the appropriate index in the result buffer which corresponds to the frame for which to store results.

MimLocatePeak1d() supports multi-frame image buffers. When passing a multi-frame image buffer, you must set MimControl() with M_FRAME_SIZE to the Y-size of each individual frame. For information on multi-frame image buffers, see the Specifying the dimensions of a multi-frame image buffer subsection of the Specifying the dimensions of a data buffer section of Chapter 21: Data buffers. Note that the total number of frames (M_NUMBER_OF_FRAMES) can be different from the frame capacity of the source image buffer. For example, if the source image buffer has a 5-frame capacity and you have 10 total frames to process, 2 calls to MimLocatePeak1d() can be done before calling MimGetResultSingle(). Typically, you would set M_FRAME_INDEX to frame 0 on the first call to MimLocatePeak1d(); then, set it to frame 5 on the second call. Finally, it is important to set M_FRAME_INDEX such that the result buffer can hold all results produced from the call. That is, there must be enough indices available in the result buffer to accommodate results from all frames in the source image buffer. For more information on peak extraction from a multi-frame image buffer, see the Extraction of peaks from a multi-frame buffer subsection of the Peak intensity detection and depth maps section of Chapter 5: Specialized image processing.

For results obtained from laser line images, you can draw an uncorrected depth map or intensity map using MimDraw() with M_DRAW_DEPTH_MAP_ROW or M_DRAW_INTENSITY_MAP_ROW, respectively. For more information, see the Generating an uncorrected depth map subsection of the Peak intensity detection and depth maps section of Chapter 5: Specialized image processing.

To verify that the peak was found at the right location, you can draw the calculated peak intensities at the calculated positions into an image buffer using MimDraw() with M_DRAW_PEAKS.