Image processing identification of laser damage induced by ns-pulsed lasers

Abstract For the purpose of quantitatively evaluating laser-induced damage in optical glass, image processing was employed. The image processing involves an improved watershed segmentation algorithm, with the aid of which different morphologies of laser damage in glass were classified. The shapes and distribution of cracks can be readily discerned in the gradient image while the severity of damage is apparent in the gray-scale image after being filtered; in addition, the marker image can be utilized to recognize the region where the damage appears; the final segmentation embodies all characteristics of damaged area and is an effective means of analyzing the laser damage in optical glass.