Image Restoration Method based on Partition and Regularization for Industrial X-ray images

Traditional regularization methods often process the whole region of images, which are suffered from some possible contradictions of detail preservation and denoising. This paper proposed an image restoration method based on partition and regularization for X-ray images of integrated circuits. Firstly, the original X-ray image is divided into detail area and smooth area based on local variance. Secondly, after filtering the smooth area, a new observed image is reconstructed by original detail area and filtered smooth area. Thirdly, Split Bregman method is adopted to solve the TV regularization denoising problem, then the image restoration results come out. Experimental results on X-ray images of LED chips and capacitors show the effectiveness of the proposed method, which is helpful to have a more accurate X-ray detection of integrated circuits.