视觉拓扑理论质疑

A Comment on the Topological Theory of Vision

对陈霖的《拓扑性质检测——计算理论的一朵可能的乌云》一文提出质疑，并提出以 下观点：视觉并非特别敏感于图形的拓扑性质：拓补性质的计算复杂度很低；视觉研究 与其他科学研究一样需要计算理论；思维科学的研究应该以形象思维作为突破口，但不 应企图离开图灵机的范畴，其研究目标首先应该是建立形象思维的计算理论。本文还提 出了一个用以判断视觉是否特别敏感于图形拓补性质的判别实验。

This paper gives a comment on the article 'The detection and mensuration of topological properties-a possible dark cloud to the computation theory' written by Chen Lin, and presents the following viewpoints: vision is not in particular sensitive to the topological properties of an image; the computation complexity of topological properties is fowl the copoputation theory is indispensable in carrying out research on vision as well as on many other disciplines; the breakthrough in the research on thinking should be focused on the image cognition and the research work should be done within the framework of Turing machine. The'object of this research is to establish the computation theory for image cognition. A discriminant experiment for deciding whether vision is in particular sensitive to the topological properties of an image is presented in this paper, too.