In this chapter, shape matching based on binary images will be discussed. The use of binary images for shape matching has advantages because morphology functions can be used prior to shape matching to modify the binary image resulting for better results.
Binary shape matching is different from geometric matching and pattern matching, which are based on edge curve extraction and pixel intensity variation in an image, respectively. Each method has advantages and disadvantages, which should be considered before selecting a particular algorithm. Binary shape matching is an efficient and effective method to determine the center of mass of matched objects from its shape. However, this method cannot find occluded objects since two objects in a binary image whose features overlap are not indistinguishable from a single object with different shape.
To use shape matching in a binary image, a color or grayscale image needs to be converted to a binary image. Prior to conversion, the objects to identify in the image should have good contrast with respect to the background so that the boundaries of each object can be accurately represented when converted to a binary image.
By using shape matching, sorting and inspection of objects are possible. The shape of the objects can be classified (sorted) and defects can be determined by comparison with a reference image (inspection).
An example of finding objects using binary shape matching is discussed in this chapter. Every ...