Абстрактный

Improvement of low illumination image enhancement algorithm based on physical mode

Shuang Niu, Yuanyuan Shang, Hui Ding, Zhong Luan

The low illumination images always have low brightness, low dynamic range and high noise, because of these characteristics, an improved enhancement algorithm is proposed for low illumination image. Considering that low illumination image enhancement will amplify the noise, it needs to reduce the noise before image enhancement. Firstly, using BM3D (block matching 3D) for image denoising in YCbCr space, then improving the transmission for image enhancement based on the physical model in HSI space. Experiments show that the method can effectively improve the brightness and contrast, enhance the detail of the image, reduce noise, and get a good visual effect.