Research On Image Processing Technology In The System Of Automatic Microscopic Chip Photo Taking

Today chip has already widely used in every aspect of work and life, and as a way to detect design, IC Reverse Engineering has attracted broad attention. The preparatory work of reverse engineering- IC image taking and stitching is the premise of reverse engineering. As the expansion of chip size, automation has become an inevitable trend of reverse engineering. In our work, we focus on two technologies in the automatic microscopic chip photo taking system: image auto-focusing and auto-stitching.In this thesis, we first introduce the principle of auto-focusing and propose an improved method of auto-focusing based on the analysis and contrast of classical arithmetic. After that, we introduce the common flow of image stitching, propose an auto-stitching methods fitting for microscopic chip images and discuss some fusion methods. Finally, we illustrate the general design of two modules: image auto-focusing and collection, image auto-stitching and fusion, and implement the function of these two modules using programming tool.The innovation in this paper is that: (1) we proposed an auto-focusing method on the basis of gray variance and Laplace transformation. This method has the advantage of gray variance and Laplace transformation, and can improve the speed and acuity degree of focusing. (2) We proposed an image stitching method based on Hough transformation. Matching images using the lines that detected in chip images can improve the speed of stitching and reduce the probability of mis-matching.