A Combined Method of Image Enhancement Based on Self-adaptive Median Filtering and Homomorphic Filtering Algorithm
GUOCHENG LIU · 2022 3rd Asia-Pacific Conference on Image Processing, Electronics and Computers · 2022
A combined method of image denoising and brightness correction based on the combination of self-adaptive median filtering and homomorphic filtering algorithm is proposed for the problems of uneven illumination and noise effects in the images of anomalous behavior study of rail traffic crowd. The combined method first uses a self-adaptive median filtering algorithm to achieve filtering and denoising processing in the null domain, and then uses a homomorphic filtering algorithm to correct the luminance of images with uneven illumination in the frequency domain, so as to improve the visual effect of the target image. The experiments show that the combined method of image enhancement can effectively achieve noise filtering of the image, making the target area of the image more even and clear, while preserving important information such as contours and edges, improving the visual effect and helping the subsequent image analysis, and providing a new image enhancement technique and method for pre-processing of image recognition in the study of anomalous behavior of rail traffic crowds.