The Detection and Recognition Algorithm of Safety Glass Fragment

Fragmentation test is one of the mandatory tests on safety glasses. We apply digital image processing techniques to the detection and recognition of glass fragments, and obtain their shape, size, and count. Based on the characteristics of glass fragment images, we propose an image segmentation algorithm based on edge detection. The first step is to detect the crack lines using edge detection and edge growth techniques. This step produces the thick crack lines from the original gray scale image. The second step then applies image segmentation based on the watershed transformation. The fragments are recognized and counted. Experiments show that the algorithm can accurately recognize the crack lines in fragmented safety glasses and determine the area of the fragments correctly. Computerized fragmentation test using our algorithm in practice can save manpower and improve test accuracy and efficiency.