The Detection and Recognition Algorithm for Safety Glass Fragment
安全玻璃碎片的检测与识别

Fragmentation test is one of the compulsory detection contents to safety glass. To apply the computer technology to the digital images which are obtained after the fragmentation test can save manpower and improve detection efficiency. Acoording to the characteristics in glass fragment images, we propose an image segmentation algorithm based on edge detection. The first step is to detect the crack lines. It includes: traditional edge detection and edge growth algorithm which can compensate the unconnected crack line regions. A thicker crack line image can be obtained which is consistent with the initial gray image. Secondly, the fragmentation particles are detected by computing the distance function and gray reconstructing and then reversing it. Finally image segmenting is based on watershed transformation and the fragmentation particles are separated by the crack line whose width is one pixel. Experimental results show that the algorithm could draw the crack line of glass fragmentation particles and divide the image into regions correctly.