视觉系统初级信息处理的时变通带滤波模型

A TIME-VARYING BANDPASS FILTERING MODEL FOR PRIMARY INFORMATION PROCESSING IN VISION SYSTEM

通过分析在视觉初级信息处理过程中时间、空间信息的依存关系，提出了随时间变化的方差σ＾２（ｔ）因子表示的时变通带高斯滤波函数模型，从数学上证明了时变带通高斯函数的微分正好实现了空间域中的滤波，由此分别讨论了零交叉检测滤波器和典型的ｏｎ－ｏｆｆ感受野的关系，计算了各特征频率通道的峰值频率，该值在一定程度上与以往的研究结果基本吻合。本文从数学上具体描述了视觉信息的提取过程是一个分时分频的提取过程的思想，

According to the analysis of spatial and temporal interrelations of the primary information processing in vision system, a time -varying bandpass gaussian filtering model which was characterized by the time-varying standard deviation σ2(t) was proposed.It was mathematically proved that differentiating of time-varying bandpass Gaussian function in frequency domain just implemented the V2G operation in spatial domain. In this opinion, the relations between zero-cross filtering and on-off receptive field were investigated. The peak frequency of each character frequency channel was also calculated from the model, and they were approximately in agreement with the experimental data.In this paper, the idea of visual information processing as a process dealing with time-varying serial spatial frequency channels is described mathematically. Thus, the united decnption of processing characteristics of visual system about spatial feature, spatial frequency feature, and time-varying property is preliminary made.