中文说明： BP神经网络用于函数拟合与模式识别，LZ复杂度反映的是一个时间序列中，高斯白噪声的生成程序，计算两个矩阵之间的欧氏距离，D-S证据理论数据融合，表示出两帧图像间各个像素点的相对情况。
English Description：
BP neural network is used for function fitting and pattern recognition. LZ complexity reflects the generation program of Gaussian white noise in a time series. Euclidean distance between two matrices is calculated. D-S evidence theory data fusion shows the relative situation of each pixel between two images.