Font Size: a A A

Research Of Coal Mine Video Monitoring System Based On Target Detection And Tracking

Posted on:2015-07-18Degree:MasterType:Thesis
Country:ChinaCandidate:G Z ZhuFull Text:PDF
GTID:2348330482982630Subject:Electrical theory and new technology
Abstract/Summary:PDF Full Text Request
China has a large number of coal mines, with the continuous expansion of the scale of mining the more and more accident caused by the special underground environment, illegal, misoperation, roof fall. To ensure safety, the enterprises are equipped with video system. But some workers in underground have little safety awareness, and the less obivious warning on the area of heavy equipment and so on, it needs further through the video monitoring system to increase supervision.According to the characteristics of coal mine video of underground, detection and tracking method is proposed for personnel target. Extraction of personnel target detectting from the video of the danger zone, and to be tracking, laid the foundation for the subsequent processing. The video monitoring system design and the software process design are given. The main work is as follows:An enhancement algorithm for video was proposed with the low illumination condition:To being Adaptive median filter in order to denosing; And then do the equal proportion transformation of each component, in order to enhance video.A fuzzy miner detection method was proposed. Some fuzzy rulers according to the characters of coal mine vedios were defined to get every pixel's membership value in the object. The silhouette and its membership function were defined base on human vison. Two functions were combined to detect the miner object.Mean-Shift combined with Kalman filter predicting algorithm is proposed, it can reduce the search time in tracking moving objects to some extent. And it also can solve the problem when the objectives speeded-up rapidly or there have a large block curtain while the Mean-Shift algorithm can't track effectively.The system is made up by server, client and the embedded devices. The server can save the video data; the client is to watch the video while request; and the last one is to collect the video. The software is in Linux environment, three parts of the software system is finished by C++. Coding and decoding is through H.264 standard.Qt is used for GUI...
Keywords/Search Tags:Object detection, Target tracking, Fuzzy theory, Embedded system
PDF Full Text Request
Related items