Detection and extraction of deficient capsule in image based on pattern-list

A method is proposed to detecting deficient capsule through image processing method. It is composed of two steps. The first step, all capsule pixels are extracted by the rectangular boxes and are stored in a rectangle box list. Then all connected rectangles are grouped to form pattern-list. Each pattern is an isolated capsule. The second step, capsule image is detected by characteristic deficient features, and picked out broken and distorted capsules. Experimental results show that not only the speed of the algorithm is very fast, but also the algorithm can be effective to the images that capsules are placed incompactly and disorderly.