点击次数：

论文类型：会议论文

发表时间：2016-01-01

收录刊物：CPCI-S

卷号：56

页面范围：26-30

关键字：Coal Bed Methane extraction; Regular reciprocating motion; Region segmentation; CodeBook; Update background

摘要：A new method of target detection based on region segmentation was proposed for the problem of reciprocating pumps and other disturbances for segmenting moving foreground. In this method, the background region is segmented by means of mean and variance, and divided into static background area and dynamic background area, and the target detection algorithm is used to identify the two different target detection algorithms. Three frame difference algorithm is used for target detection in the static background area, and the interference of the reciprocating motion of the pump is eliminated. In the dynamic background region, an improved codebook algorithm based on gray level image was used to target recognition. The simulation results show that the method has very good practicability, which greatly reduces the false detection rate, and the algorithm meets the requirement of accurate target identification in the mining field of coalbed gas in real-time, accuracy etc.