Image Vignetting Correction Based On The Segmented Radial Gradient Symmetry

Infrared imaging can convert infrared radiation information into visible light information that human vision can perceive.It is widely used in the military and civilian fields.In recent years,with the continuous development of infrared detectors in the direction of high resolution and high frame rate,there comes some new problems of infrared image quality degradation in the detectors.The bright-vignetting noise is one of these problems.The bright-vignetting noise will lead to the loss of the detailed information of the image,which hinders the application of infrared detectors in various fields.How to correct the bright-vignetting noise in the infrared images is a problem needed to be solved in the field of infrared image processing.In order to correct the bright-vignetting noise,this paper researches the cause and the characteristic of the bright-vignetting noise.By collecting the original infrared image data,we find that the cooling capacity between the central area and the peripheral area of the focal plane is quite different.Therefore,the heat in the central area is higher than the heat in the peripheral area of the focal plane,which results in the inconsistencies of the radiation response between the central and peripheral areas of the focal plane,and ultimately leads to the appearance of the bright-vignetting noise.Besides,by collecting the radial noise factor of the infrared focal plane,it is found that the energy distribution of the bright-vignetting noise has a segmented characteristic in the radial direction.Considering on the characteristic of the bright-vignetting noise,this paper proposes a vignnetting correction algorithm based on segmented radial gradient symmetry.Based on the optimization function of the original radial gradient symmetry optimization algorithm,the algorithm adds a fidelity function term to correct the brightvignetting noise in the premise of retaining the original image information as much as possible and avoiding the over-correction of the vignetting nosie in the corrected image.Besides,the algorithm proposes a segmented radial correction function to correct the vignetting noise according to the characteristics of bright-vignetting noise so that the correct function can fit the bright-vignetting noise better.The results of simulation experiments and scene-based experiments show that the proposed algorithm can correct the bright-vignetting noise effectively.The PSNR of the corrected image is increased by 27.6%,the SSIM is increased by 2.6% and the energy gradient function is increased by 11.1%.