Research On Enhancement Algorithm Of Low Illumination Image

In recent years,video surveillance systems have become more and more developed,but in low-light environments,due to insufficient illumination,the image brightness is low,the detail information is lost seriously,and the recognition of images is difficult.Therefore,the research on low illumination image enhancement algorithms is important theoretical significance and application value.In this paper,the similarities between low-intensity image inversion images and foggy images are studied.For the dark primary color prior defogging algorithm,the image encapsulation and the time of the soft-spot method used for the transmittance refinement will appear.For the problem of high complexity,this paper proposes a low-light image enhancement algorithm with improved dark primary color prior and histogram equalization.The constant is used to replace the minimum filter,and the image is processed pixel by pixel to quickly solve the reduced transmittance.By introducing parameters to compensate for the reduced transmittance,image distortion problems caused by large channel minimums of image pixels are prevented.The ambient light value is estimated using a quadtree algorithm.Then,the histogram equalization processing is performed on the image processed by the dark primary color prior algorithm,so that the image gray histogram is transformed from a non-uniform distribution to a uniform distribution in the gray scale range,further improving the contrast,and finally performing image inversion.An enhanced low illumination image is obtained.The simulation results show that the improved algorithm improves the brightness and contrast of the low-illumination image,effectively eliminates the block effect,and the image color reproduction is high,which improves the visibility of the image,and the running speed is obviously improved compared with the soft-spot method.However,the enhancement effect on the depth of field of low illumination images is not obvious.The shortcomings of the above algorithms and the traditional homomorphic filtering algorithm transfer function parameters are too difficult to control and the multi-scale Retinex algorithm is prone to halo artifacts,an improved low-light image enhancement algorithm is proposed.Firstly,the R,G and B components of the low illumination image are enhanced by the improved homomorphic filtering algorithm to improve the image contrast;then the image is converted from the RGB color space to the HSV color space;the saturation component is in the HSV color space.Adaptive nonlinear stretching is performed,and the enhanced multi-scale Retinex algorithm is used to enhance the brightness,the illuminating component is corrected by gamma conversion,the reflected component is processed by Sigmoid function,and finally convert the image to RGB color space output The image is simulated by Matlab.Experiments show that the algorithm improves the information entropy,peak signal-to-noise ratio and contrast of low illumination images,and improves the visual effect of low-illumination images.