An Improved Vibe Algorithm to Detect Personnel Underground in Coal Mines

Abstract

Coal mine safety has always been the most important prerequisite for underground coal mine work. Mine personnel inspection is an effective means to ensure underground safety production. Therefore, the quality of inspection will play a decisive role in safety production. At present, due to the influence of the complex environment in coal mines, ghost images are prone to appear in the process of personnel detection, which has a certain impact on the accuracy of detection. Aiming at this phenomenon, a Vibe method for secondary detection based on ghost images is proposed. In the process of underground coal mine personnel detection, the minimum bounding rectangle of the personnel area is delineated, and each pixel of the personnel area and all the areas outside the rectangle are calculated separately. The process of judging whether it is a ghost image and eliminating the ghost image by the number of pixels whose similarity reaches the threshold. Through subjective and objective verification, the proposed improved algorithm has been effectively improved compared to the traditional Vibe algorithm and the Vibe+ algorithm, which is prone to ghosting problems. In terms of the accuracy, recall rate, F1 value and other objective evaluation indicators of the algorithm model, it is proposed Compared with the two algorithms, the improved Vibe algorithm improves by 2.71%, 4.79%, and 3.73% respectively. Experimental data shows that the improved Vibe algorithm effectively suppresses the appearance of ghosts in the process of underground coal mine personnel detection, improves the accuracy of foreground and background separation, enhances the ability to detect moving targets in coal mines, and provides technical support for safe production in coal mines.