关于处理电视信号(如电视电话或会议电讯)的频带压缩或信息冗余度的压缩问题已有很多方法。【1-9】本文提出另一种处理方法。对于一个活动图象所需传送的信息仅是各帧信号之间的差一帧差信号,这一部分正好是图象的活动部分。由于人眼对于图象的活动部分的分辨力是不高的,所以我们可以传送帧差信号的近似值。为了得到帧差信号的线性近似,本文利用Kalman滤波。由于人眼在高亮度下可容许更大的亮度误差。因而Kalman滤波器被设计得在高亮度信号下比低亮度时容许更大的误差。这样在信号编码过程中可引入更大的容差,根据码率—容差理论,信息量可得到进一步压缩。为了对帧差信号的线性近似进行编码,我们只需对它们的端点进行编码,这样所传送的信息就仅为各端点的幅变和位置,进一步考虑到各行线段端点间的相关性,更多的信息冗余度可以被压缩。通过计算机模拟,证明了用自适应Kalman滤波线性近似来传送帧差讯号可大量压缩信息冗余度。例如,处理一个移动的手,对它图象编码的信息量约为0.3—0.4bit/pel. There are many ways to deal with the compression of bandwidth or the redundancy of information in television signals such as videophones or conference telecoms. [1-9] This article proposes another approach. The information to be transmitted for a moving picture is only the difference of one frame of difference between the signals of the frames, which is exactly the active part of the picture. Since the eye's resolution of the active part of the image is not high, we can send an approximation of the frame difference signal. In order to obtain a linear approximation of the frame difference signal, this paper uses Kalman filtering. Due to the high brightness of the human eye can allow greater brightness error. Thus, the Kalman filter is designed to tolerate more errors at high brightness than at low brightness. This can lead to a larger tolerance in the signal encoding process, which can be further compressed according to the rate-tolerance theory. In order to encode the linear approximation of frame difference signals, we only have to encode their endpoints such that the information transmitted is only the amplitude and position of each endpoint, further taking into account the correlation between the endpoints of the line segments, and more The redundancy of information can be compressed. Through computer simulation, it is proved that using the adaptive Kalman filter linear approximation to transmit the frame difference signal can greatly reduce the information redundancy. For example, dealing with a moving hand, the amount of information encoded in it is about 0.3-0.4 bit / pel.