Research on submicron measurement based on computer micro-vision

To overcome hampers in applying computer micro-vision technology on submicron scale measurement, a new image enhancement algorithm is proposed. Combined with image deblur and wiener filter, as well as the snake model counter extraction, the measurement is well developed. Based on illumination-reflectance model and Center/Surround method, the Gaussian scale parameter is reasonably selected through maximizing the contrast average and minimizing the contrast variance among several windows. The enhancement algorithm can automatically balance the dynamic range compression and image information retaining. Experiments show that the measuring method can clearly observe the submicron scale small to 833nm and achieve little relative error for micron