The Study On Radiographic Image Enhancement Based On Total Variation-Based L₀ Minimization

According to the visualization of the spatial position and direction of the actual physical structure,not only makes it possible to detect,but also helps to understanding the basic mechanism of the defect development and device failure.There are several kinds of commonly methods in NDT and NDE,such as X ray imaging,acoustic imaging and impedance imaging.Each method has its important application and limitations.Most of the methods for empty and large aperture are sensitive.Depending on the type of radiation,ray can be divided into two kinds,the particle radiation and electromagnetic radiation.The particle radiation includes the neutrons,protons,alpha particles,beta particles and so on.The electromagnetic radiation includes X-rays,gamma rays and so on.Neutron imaging technology is a powerful material nondestructive testing tools,and is the only radiography imaging technique that can detect materials?such as nuclear waste material?under the background of high gamma radiation,and has been widely used in industrial and materials research.Radiography imaging technique can show the internal structure of the tested object in non-destructive way by image.It plays an important role in the non-destructive testing of security,medical,industrial flaw detection.Low contrast and high dynamic range is inevitable for the digital radiography images due to the property of the radiography imaging system hardware.Image quality is difficult to meet the prescribed standards for industrial and medical applications.For this situation,introduce the ?₀ fidelity term to rebuild the conventional models and develop a new model of constrained ?₀ fidelity variation model?TV-l₀?.Combined with image layer separation for dynamic range adjustment in the basic layer and detail enhancement in detail layer.Furthermore,by applying alternating direction method of multipliers?ADMM?to solve the model,we derive fast convergent iterative algorithm which is applicable for detail enhancement.The numerical experiment results show that TV-l₀ algorithm can effectively enhance contrast,image details,remove noise and improve image visual effect compared to existing algorithms.TV-l₀ algorithm has universal applicability for radiographic image enhancement and is important to improve practical significance for nondestructive testing technology in low dose radiation.It has a very goodapplication prospect in industrial nondestructive testing and medical detection.