IC Defect Image Filtering Based on Graph Morphology

Integrated circuit(IC) defect image filtering is very important for the classification and the identification of IC real defects and its result will affect the validity and reliability of the later processing and analysis. This paper aims to use the new theory of graph morphology to filter the IC defect image. In this paper, we introduce the morphological operations on graph. The image is represented as graph composed of vertices, edges connecting vertices and weight. Structuring graph which is a counterpart of structuring elements in classical morphology is presented. We use the graph morphological filter composed of opening and closing to realize filtering. Experiments show that the application has better result compared with traditional methods for filtering the images.