9Geometric Matching

Pattern matching algorithm described in Chapter 5 uses the pixel intensity information as the primary feature for matching. As an alternative, geometric matching uses boundary edges to characterize the shape of an object and then uses this characterization to search for similar shapes. To use this method, the object and background should be distinguishable by sharply contrasting regions in order to accurately determine the boundary of the object. The boundary shape information of the objects is compared with that of an object in a template image to determine similarity. If the edge of the boundary is not sharp, pattern matching as described in Chapter 5 is recommended. An advantage of geometric matching is that it can find matching objects regardless of shifting, rotating, scaling, and even occlusion (overlapping of objects in the image). Geometric matching can be used in the following applications: gauging, inspection, alignment, and sorting.

Example VI on geometric matching can be found from the following folder:

Figure 9.1 shows example VI provided with LabVIEW. As seen in the Template image in Figure 9.1, the object has a distinct boundary to extract the geometric features. By using the appropriate search parameters, the rotation angle, scale, and the location of the matched patterns can be determined. Note that the method detects boundary ...

Get Practical Guide to Machine Vision Software: An Introduction with LabVIEW now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.