Research Of Industrial X-ray Image Enhancement Algorithm

X-ray imaging technology is an advanced nondestructive testing technology,which can reproduce the internal characteristics and identify the material information completely without damaging the overall structure of the object.At present,the technology has been widely used in medical diagnosis,security inspection,industry and other fields.However,industrial X-ray detection images usually contain a lot of background noise,poor contrast and image blur,these adverse factors to industrial X-ray detection and X-ray image analysis caused great difficulties.In order to improve quality of image and enhance the details,this topic from two angles of histogram equalization and Retinex theory study,puts forward the improved histogram equalization and bilateral filtering of X-ray image enhancement algorithm and X-ray image enhancement algorithm based on multilayer model.The main work of this paper is as follows:Based on histogram equalization and bilateral filtering of X-ray image enhancement algorithm contains three stages,the first stage is the improved histogram equalization algorithm,the algorithm set high and low threshold,the histogram is divided into three sub histogram and compute the probability density function,cumulative distribution function to modify the output function curve.Secondly,bilateral filtering is used to denoise image.Finally,the global huemapping method is adopted to compress the dynamic range of the image,and the image can be clearly displayed on the device.Through Result analysis,The algorithm can effectively improve the contrast of the picture,suppress the noise and has more details,so that the processed image has good visual effect.X-ray image enhancement algorithm based on multi-layer model.The algorithm uses Stype function which can not only stretch the pixel level of the image,but also protect the boundary pixel value to replace the logarithmic function of the traditional Retinex theory,In the process of decomposing the image into illumination component and reflection component,the guided filter is introduced to optimize the illumination component.Then with double logarithmic transformation to illuminate component processing,next to illuminate component will continue to split and optimize,until meet the termination conditions,finally keep brightness of light weight,the reflection components of each layer,and finally a layer of light weight fusion.Through Result analysis,in terms of visual effect,the algorithm can avoid artifacts.It also has great advantages in quality evaluation.