Automated Defect Analysis of Material Surface Defects

Analysis of material surface defect plays an important role in the development of new materials and fabrication processes, and in the diagnosis of process problems in material manufacture. The task of manual defect analysis is labor intensive and tedious, and the level of automation available is limited. We propose to develop image processing techniques and a complete system to fully automate defect analysis. Features that quantify size, shape, orientation, and distribution of defects will be selected according to user needs. Algorithms for efficient computation of these defect features will be developed. Capability for graphic data display will be incorporated.