Research On Key Technologies Of Super-resolution Restoration For Underground Mine Surveillance Images

Super-resolution restoration is a signal processing method. It uses the prior knowledge of images or complementary information of low resolution image sequence to restore the image information which is out of cut-off frequency. Super-resolution restoration can improve image quality and obtain higher resolution image. Due to the influence of production environment, production cost and other factors, the underground mine surveillance images are fuzzy, contain a lot of noise. And the resolution of images is low. This paper discusses the development status of super-resolution restoration and analyses the characteristics of the underground mine surveillance images. The key technologies of super-resolution restoration for underground mine surveillance images are researched by using of the theory of partial differential equation, non local means, fuzzy set and Markov random field. This paper puts forward four methods to realize the super-resolution restoration of underground mine surveillance images. The restored images can show the production of coal mine clearly, completely and accurately. The results can meet the needs of coal mine monitoring.