Paper Title
Research on Coal Mine Safety Monitoring Action Recognition System based on Kernel Function Method

Abstract
This paper proposed an application solution for the intelligent identification system of coal mine safety monitoring action based on kernel function. We used the object detection model based on kernel density estimation for dynamic background segmentation, combined the adaptive update strategy to extract the foreground moving object. We implemented a method of action feature classification based on SVM model based on mixed kernel function and optimized the model parameters. This paper achieved the coal mine safety monitoring intelligent recognition system framework, provided a solution for intelligent mine monitoring information management, improved the efficiency of coal mining intelligent monitoring management. We could find the hidden danger of safety in coal mine production effectively and timely, and reduced the probability of the occurrence of coal mine safety accidents. Keywords - Coal mine safety, Intelligent monitoring, Behavior recognition, kernel function; SVM.