Research On Image Enhancement Algorithm In Scotopic Vision Environment

In the scotopic vision environment,the image acquisition sensor collected image has the problems of low brightness,detailed information,outstanding noise,low contrast,which would seriously affect the vision of the human eye recognition ability,and subsequent image processing.In order to solve these problems,this thesis improved an image enhancement method,which enhances the low-quality images imaged in the scotopic vision environment.Thereby human eyes and computers can recognize the real information hidden in the images.Aiming at the problem that human eyes cannot recognize image information due to the low brightness of image in scotopic vision environment.this thesis improved an image brightness enhancement method based on Retinex in scotopic vision environment.First,estimate the global illumination image and the local illumination image of the image in the scotopic vision environment,and weight the average pixels of the global illumination image and the local illumination image to obtain the preliminary estimated illumination image with global and local characteristics;Secondly,the improved weighted guided filter is used to perform multi-scale edge-preserving smoothly filtering on the initially estimated illuminance image to obtain the filtered illuminance image.Then adaptive gamma correction is performed on the filtered illuminance image to obtain the corrected illuminance image.Finally,according to Retinex theory and the corrected illuminance image,the reflection image was reflected the detailed information of the image is obtained,that is,the image in scotopic vision environment with enhanced brightness.Aiming at the problems that the image detail information is not prominent,the noise is large,and the contrast is low in the scotopic vision environment,this thesis carries on the image detail enhancement processing on the basis of the image brightness enhancement in the scotopic vision environment.Firstly,block-matching and 3D filtering denoising process is performed on the brightness channel image of the reflected image,and the color channel of the reflected image remains unchanged.Secondly,the image contrast enhancement process is performed on the denoised brightness channel using the unsharp masking method.Finally,the contrast enhanced brightness channel image and the reflected image color channel image are recombined into an image with enhanced detail.In order to verify the effectiveness of the image enhancement method in scotopic vision environment proposed,an image simulation and comparison experiment was conducted,and design subjective evaluation experiments.Experimental results show that this method in this thesis can effectively enhance the brightness,contrast and sharpness of the image in scotopic vision environment,and it has well results both subjectively and objectively.