Cold Reflection Suppression And Enhanced Display Technology Of High Dynamic Range Infrared Images

High dynamic range images,compared with ordinary ones,could not only provide a larger dynamic range,but also a better reflection to the real subject and scene.The high dynamic range infrared image is taken by the high-performance infrared image detector under high dynamic infrared scene,whose output signal usually ranges 14 or 16 bits.However,due to the influence of detector materials and manufacturing technology,obvious defects in high dynamic range infrared images are caused inevitably,such as low contrast,fuzzy edge,low signal-to-noise ratio and low spatial resolution.In addition,the cold reflection is also a significant defect in the infrared image.Characteristics mentioned above will affect the visibility of infrared image,and furthermore influence the subsequent processing and application of infrared image target detection,recognition and tracking.Therefore,the purpose of this paper is to explore infrared image optimization and dynamic range compression display with high dynamic range information from images.Based on the purpose,the following work is conducted:First of all,aiming at the cold reflection phenomenon in infrared image,we investigate causes of this image defect and the existing solutions,and on the basis,analysis on the specific manifestations of the cold reflection phenomenon in the spatial domain is made.Further,we have interpreted the problem as the non-uniformity of low-frequency information in the frequency domain.Finally,based on the above understanding,from the perspective of digital image processing,two different solutions are proposed and mainly elaborated:the cold reflection suppression algorithm based on wavelet transform and the cold reflection suppression method based on the simulation of the cold reflection intensity model.The experimental simulation shows that the two algorithms are able to effectively suppress the cold reflection phenomenon in the infrared image.After comparing the average operation speed of the two algorithms,the cold reflection suppression algorithm based on wavelet transform is selected as a part of the proposed cold reflection suppression and enhancement display algorithm of infrared image.Secondly,as a supplementary processing of the cold reflection suppression algorithm and in response to the demand of improving the visibility of infrared image details,this paper mainly explores the details enhancement method in wavelet domain.Under the premise of making full use of the advantages of multi-resolution of wavelet decomposition and the characteristics of noise intensity weakening in wavelet domain,the paper finally proposes and concretely expounds the infrared image detail enhancement algorithm based on the combination of wavelet transform and Laplace operator combining with the knowledge of edge detection operator and mask algorithm in spatial domain,and verifies its effectiveness.Then,following the processing of the proposed cold reflection suppression algorithm in this paper,an inevitable concentration trend of infrared image contrast is presented,resulting in further decline of the processed image contrast,which appeals to improvement of the contrast after above two steps.Based on the study of several contrast enhancement and mapping algorithms,this paper first highlights the advantages of Log-based Histogram Modification algorithm in the mapping display of high dynamic range infrared image,and then proposes a further improvement method to expand the advantages of 1 Log-based Histogram Modification algorithm in the visibility of details of infrared image among the bright and dark areas.Finally,based on the simulation of the proposed algorithm in Matlab environment as well as the classical algorithm,firstly,the paper analyzes whether our algorithm could obtain the infrared image that is human eyes-friendly by testing images in the test image set.And then we use the necessary image evaluation indexes to comprehensively compare and evaluate the image processing effect according to the two subjective physical quantities of contrast and detail richness.The results show that the algorithm proposed in this paper is feasible and superior in cold reflection suppression and enhanced display.