Research On Pseudo Color Enhancement Of Polarizing Image

Polarization imaging technology realizes the separation of target and background to some extent based on the difference of the polarization characteristics of objects,and improves the contrast of images.But the resolution of the human eye to the gray level is far below the color resolution.In order to improve the detection and recognition rate of the target in the polarizing image,this thesis studies the pseudo color enhancement method based on the polarization parameter combination and the pseudo color enhancement method based on the polarization parameter fusion.The main contents are as follows:(1)The pseudo color method commonly used in ordinary images is studied.Taking the scene one image obtained in this paper as an example,the MATLAB is used to simulate and show the advantages and disadvantages of the algorithms.(2)According to the characteristics of each polarization parameter,we choose suitable color gradation mode and coding method,and use pseudo gray color transform to make pseudo color enhancement.Then the threshold of the above method is redivided by the maximum inter class variance method.The enhancement effect of the pseudo color image is obvious after the threshold is divided,and the different objects in the scene can be quickly resolved by color.(3)A pseudo color enhancement method based on polarization parameter combination and multi threshold division is proposed to make the image contain more information.The polarization characteristic factor is derived based on the transmission law of polarized light in the medium,and the intensity image is used to modulate the characteristic factor image.Based on the polarization image,we divide the threshold by the iterative method and the maximum variance method,and finally enhance the pseudo color.The enhanced pseudo color image contains abundant information,which effectively improves the detection and recognition rate of targets in complex background.(4)In the study of pseudo color method based on polarization parameter fusion,a pseudo color fusion method based on NSCT transform and HSI model is proposed.First,the polarization and intensity images are fused by NSCT transform to get the polarization feature image,and then the intensity image is enhanced with the average logarithmic natural brightness value.The processed polarization feature image and polarization angle image are mapped to the H channel according to certain mapping rules.Then according to the characteristics of polarization angle and polarization feature image,the value of S channel component is obtained by nonlinear function formula.The enhanced intensity image is mapped to the I channel,and the final pseudo color enhancement image is obtained.The results show that the algorithm improves the overall visual perception of the image compared with other algorithms,and the details of the image sharpness and the outline texture are also improved.