摘要
运动目标轮廓识别是提升煤矿井下监控预测价值的基础,也是监控视频系统的开发难点。通过提出PCA-SIFT算法,运用该算法对煤矿监控运动目标进行识别,并将识别结果与传统Mean Shift算法对比。结果表明:PCA-SIFT算法可更加清晰地识别出井下图像轮廓,其帧处理效率和正确率更高,且运动目标跟踪误差十分稳定,可有效防止跟踪目标丢失。
Moving object contour recognition is the basis of improving the value of monitoring and forecasting in coal mine, and it is also the difficulty of monitoring video system. Through the proposed PCA-SIFT algorithm, the algorithm is used to identify the moving target of coal mine monitoring, and the recognition results are compared with the traditional Mean Shift algorithm. The results show that the PCA-SIFT algorithm can be used to identify the image contour more clearly, the frame processing efficiency and accuracy are higher, and the tracking error of the moving object is very stable, which can effectively prevent the tracking target from losing.
出处
《煤炭技术》
北大核心
2017年第12期230-232,共3页
Coal Technology
