Research On Video Stitching Algorithm Of Coal Mine Face

The environment of the coal mine working face is complex,and effective video monitoring can ensure safe production underground.But the existing underground camera observation area is limited,and the multiple monitoring cameras composed of multiple cameras have many overlapping screen images,which is likely to cause low video utilization and affect relevant personnel.Video stitching can effectively solve this problem.However,the related algorithms in the current video stitching technology have adaptability and poor real-time performance,so it is of great significance to study the video stitching algorithm for the underground working environment.In order to make full use of the effective information of surveillance video from multiple cameras,the redundancy of surveillance video is removed.In combination with the practical situation of the coal mine,in view of the existing video Mosaic algorithm in the application of repeat similar video frame image features extraction algorithm more iterations and splicing speed to solve the problem of poor cannot meet the demand of application in the image matching technology based on the proposed a fast video mosaics of splicing method,the effective of downhole monitoring video stitching operation.By analyzing common stitching algorithms and based on SIFT feature extraction,an optimized image stitching algorithm was proposed.Moravec corner detection and PCA principal component analysis were adopted to optimize the feature extraction process.The feature approximate nearest neighbor matching algorithm and the improved Ransac algorithm are used to filter matching feature points,calculate the transformation matrix,and use the fade-in-out fusion method to realize the image stitching operation.The optimized image stitching algorithm combined with the video interframe stitching method is used to stitch the downhole video,and the perceived hash method is used to determine whether the collected video frames are similar,to avoid repeated extraction of image features of similar video frames,and to further improve the efficiency of the algorithm.After stitching downhole video,based on the initial frame stitching template,the optimized fast video stitching algorithm based on perceptual hashing method can stitch 23 frames per second on average.The experimental results show that the video stitching method proposed in this paper can effectively meet the real-time requirements,and has certain stability and applicability in the splicing effect of downhole working face.