An Adaptive Detection Method for Early Smoke of Coal Mine Fire Based on Local Features

Aiming at the particularity of coal mine conditions, and in order to improve the effectiveness of smoke detection method in actual early detection, an early fire image smoke detection method based on adaptive histogram algorithm is proposed in this paper. Firstly, the principle of adaptive histogram algorithm for smoke image is explained, and the calculation method of line by line sliding window histogram is introduced. Then, the smoke detection process is given based on image adaptive histogram feature to realize coal mine smoke detection. Finally, the method is used to realize target detection in different smoke scenes. In the experiment, the traditional edge detection operator, smoke detection based on global histogram and the adaptive histogram smoke detection in this paper are used to detect smoke image and flame smoke respectively. The results show that this method can better reflect the motion characteristics of smoke at the edge, and contains large information entropy, which shows that this method shows good performance in smoke detection.