Research On Industrial Radiographic Image Enhancement Algorithm

In recent years,the advancement of X-ray imaging technology has made it a very effective non-destructive testing method.It can clearly and accurately display the internal structure and defects of objects,and has been widely used in safety inspections and medical diagnosis.However,in actual production,the complex structure of the tested object,uneven thickness changes,and information distortion in the image acquisition and transmission of the imaging equipment can lead to low contrast in industrial X-ray images and unclear display of small details.Therefore,in order to meet the practical application needs of non-destructive testing,it is necessary to improve the image quality,not only to improve the overall contrast but also to preserve the detail information in the original image,so that the key information in the image can be clearly displayed.Industrial X-ray image enhancement technology has made significant progress.This article explores different algorithms in this area and evaluates their performance through experimental data.Moreover,by combining the advantages and disadvantages of several different algorithms,this article proposes two algorithms for enhancing industrial X-ray images.The specific research content is divided into three parts:1.This article will explore the development history of industrial X-ray detection systems and discuss how to use digital image enhancement technology to improve the processing effect of X-ray images.Two related digital image enhancement processing technologies,high dynamic image tone mapping technology and low illumination image enhancement technology,are studied for the high dynamic and low illumination characteristics of industrial X-ray images.2.Based on the high dynamic and low illumination characteristics of X-ray images,the image enhancement is decomposed into two sub-problems: dynamic range compression and low illumination image enhancement.Firstly,the high dynamic range tone mapping algorithm is used to compress the image,and then the normalized gray value processing is applied to the compressed image.The gradient field reconstruction technique is used to restore the image,and then the low illumination image enhancement technology is used to further increase the image contrast,eliminate useless information such as background,and then the unsharp mask operator and restricted adaptive histogram equalization technology are used to further enhance the image details.Through comparative experiments,the contrast of various parts of the workpiece is significantly improved,the edges are naturally transitioned,and the details and defect parts are clearly displayed.3.Based on the characteristics of industrial X-ray images,the image multi-scale decomposition is improved and combined with the local edge-preserving filter algorithm.Firstly,the original high dynamic industrial X-ray image is logarithmically compressed to compress the dynamic range.Then,the basic layer is obtained by processing with the local edge-preserving filter,and the detail layer representing the small gradient is obtained by subtracting the logarithmically compressed image from the basic layer.The edge-preserving filter is iteratively applied to the basic layer with increasing filtering radius to obtain multiple detail layers,which are mapped by S-curve and fused,then added to the structural layer linearly compressed and gamma-transformed,then normalized.The resulting image is processed by contrast-limited adaptive histogram equalization to obtain a low dynamic range image with high contrast between the background and the workpiece and clear details.The experimental results prove that the proposed method has a significant enhancement effect on complex structures of workpieces and high algorithm efficiency.