Image Correction In Security Inspection Equipment

Security inspection equipment has become the necessary device for the public safety. The quality of image is an important evaluate criterion of security inspection equipment. The main purpose of this thesis is to analyze and solve two problems that decline the quality of image, the non-uniformity of image and image blurring.To solve the non-uniformity of image, the causes of this problem are analyzed. Based on this analysis, a threshold T is given to divide image gray into high intensity area and low intensity area. The causes of non-uniformity of this two areas are different. For high intensity area, the response model of detector is corrected. A higher order term is added to the response model. Based on this model, a corrected two-point correction algorithm is designed. The experimental results of this algorithm are compared to the results of two-point correction algorithm. And these comparisons verify the effectiveness of this algorithm. For the low intensity area, a compensation correction algorithm is investigated. The contrast of images before and after correction verifies the effectiveness of the algorithm. Finally, the selection of threshold T is introduced. Then the corrected two-point algorithm and compensation correction algorithm are used together to correct the image. The experimental results show that the quality of image is improved.To solve the problem of image blurring, the reasons of blurring are discussed and the model of degradation are developed. Then, some classical restoration algorithms and regularization restoration algorithms as well as their advantages and disadvantages are discussed. Based on these works, total variation image restoration algorithm is adopted. The experimental results show the effectiveness of this restoration algorithm.