Research On Industrial-ray Image Detail Enhancement Algorithm

X-ray detection technology is one technology of non-destructive testing. It plays animportant role in industrial inspection, security and so on. However, due to the variouselements influence on hardware of the X-ray detection system as well as the size and nature ofthe checked work-piece. It will result in reduced quality of the ray image and visual effect ofvariation. For example, the image is blurred, contrast reduction, edge and detail loss, etc. Thiscauses some difficulties in distinguished defects of ray images. Therefore, the need for imageenhancement processing to meet industry needs.This paper describes the status of ray image detail enhancement at first. Followed by itstudy of the industry-ray image detail enhancement theory and technology in-depth andanalyzes the shortcomings of some existing algorithms. Based on these, it put forward someimproved image detail enhancement algorithm:1. Studied the most basic top-hat transform operator and an improved new top-hattransform operator of the morphological. And the combination of this new top-hat transformoperator and toggle contrast operator. It proposed a new algorithm. The algorithm is appliedto ray images. It can get better processing effect than the original algorithm.2. Studied two improved generalized unsharp mask algorithms. The first algorithm iscombined tangent operator, improved median filter operator and contrast enhanced operatorbased homogeneity to improve industrial-ray image detail enhancement methods. The secondalgorithm is combined fuzzy operator, iterative median filter and top-hat transform to improveindustrial-ray image detail enhancement methods.3. Studied two kinds adaptive image enhancement algorithm based Teager operator. Thefirst algorithm is adopted the Teager operator taking the expansion of vertical direction,horizontal direction and a diagonal direction into account at the same time. And enhanceimage detail based on image gray entropy adaptive. The second algorithm is adopted the Teager operator can adjust the ratio of smoothing and sharpening. It is possible to effectivelyenhance the image and suppress noise enlarged. It is based on the ratio of the neighborhoodstandard deviation and mean value to enhance image on the adaptive gain respect.4. Studied several improved algorithm based on human visual characteristics and localvariance. Because the edge region variance of image higher. And the human eye isparticularly sensitive to changes in gray value of image median area. In this area should beused in small gain value. Therefore, the adaptive gain function is set to image variance andthe value which is the subtraction between the pixel value and the median value relatedfunctions.