中文说明：输入视频转换为frames.RGB视频，然后转换到HSV model.In为了提取前景，我们执行背景减除算法是基于高斯混合模型。我们计算的平均值和对于每一帧的标准偏差，并与预定的阈值进行比较在连续两帧之间的平均值的差异。如果该值大于它被作为其他它被认为是作为背景的前景的阈值。然后，我们使用Canny算子来改善前景的边界的质量进行边缘检测。为了跟踪前景和把一个边界框，我们使用了卡尔曼滤波。
English Description：
The input video is converted toframes.RGB video is then converted to HSV model.Inorder to extract the foreground, we perform background subtraction algorithmwhich is based on Gaussian mixture model. We calculate the mean and standarddeviation for every frame and the difference in the mean between twoconsecutive frames is compared with the predefined threshold. If the value isgreater than the threshold it is taken as the foreground else it is consideredas the background. Then we perform edge detection using canny operator toimprove the quality of the boundary of foreground. In order to track theforeground and put a bounding box we use the kalman filter.