欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对摄像机静止的情况，提出了一种可运用于实时监控中的运动目标检测与跟踪的方法、采用更新函数实现背景实时更新，通过差分算法检测运动目标．在跟踪模块中，提出建立帧间目标“关系矩阵”实现多个运动目标匹配，并采用卡尔曼滤波器预测目标参数，在运动目标相互遮挡的情况下，根据预测参数跟踪目标，获得目标轨迹、通过多个图像序列测试，算法具有良好的实时性和适应环境变化的能力．
英文摘要：
This paper presents a method of moving object detecting and tracking under stationary camera, which can be used for real-time surveillance. The background was updated by a updating function, and then, moving objects were detected by difference. In the tracking module, a ＂relation matrix＂ was built to realize multiple objects recognition, and the parameters of objects could be predicted by using kalman filter for tracking, and then, locus could be obtained even when a dynamic occlusion occurred between two or more objects. This method has been tested on image sequences to show the validity of the approach.
同期刊论文项目
同项目期刊论文
期刊信息
位置：