An Inage Denoising Algorithm Based On Visual Receptive Field Mechanism

Image is an important way for people to obtain information, however, when we got the image,it inevitably be mixed with noise,and the noise would take a lot of problems for subsequent image processing, so the image denoising is a very important area of research in the image processing. So far, researchers have proposed many methods of image denoising, There are classical Gaussian filter, bilateral filters. It’s advantages are convenient calculation method, the idea is simple. The main drawbacks are blurring the image, high-frequency noise can not be completely removed. We know that the human brain is the most effective known biological intelligence system, the human visual system to image recognition, processing and the ability to handle far more than any of the existing computer and information processing systems. However with the mechanisms of neural information science are applied to computer vision, it brought vitality for computer vision research. In this paper,we constructed image denoising algorithm based on the receptive field model(BRF), and try to find a new way for image denoising research.We constructed image denoising algorithm based on the receptive field model(BRF), and achieve the purpose of image denoising successfully. The algorithm consists of two parts, ‘judge noise’ and ‘noise processing’, the two parts realized by the corresponding receptive fields, and the judge noise is the basis of noise processing. After noise image computed by judege noise algorithm, we can know the size and location of the noise, which is then treated with a noise process algorithm. we used ON and OFF type receptive field adaptive suppression noise seperately,to achieve the purpose of image de-noising lastly.The experimental results show that the algorithm can not only suppress noise while preserving image detail. After contrast with the classical iamge denoising algorithm,such as Gaussian filter, bilateral filter, as well as popular non-local means filter, BM3 D, total variation. it is found that the effect of denoising is very good.The visual mechanism applied to the image denoising, not only achieved satisfactory results, but also open a new door for image denoising research.