基于早期视觉的算法模型评述  

RESEARCH BASED ON EARLY VISION ALGORITHM MODEL

作　　者：方习之[1] 危辉[2] 

机构地区：[1]复旦大学计算机科学系认知算法模型实验室,上海200433 [2]复旦大学脑科学研究中心,上海200433

出　　处：《计算机应用与软件》2007年第11期14-15,共2页Computer Applications and Software

基　　金：国家重点基础研究发展规划973项目(2001CB309401);国家自然科学基金项目(60303007)的资助

摘　　要：计算机视觉与神经生理学相结合,早期视觉模型与计算相结合,对视觉研究领域具有深远的意义。早期视觉研究处理的视觉信息是视觉系统接受信息的第一步,是视觉研究的基础。它用计算机视觉的实现方式,模拟了视锥通路的早期视觉信息处理模型,验证了生理视网膜在信息处理过程中的特性,分析了视觉神经在信息采集必须充分必要的条件下,硬件复杂度、处理时间与处理精度之间的平衡。The synergy between computer vision and neurophysiology as well as early vision model and algorithm has far-reaching significance to vision research field. The vision information processed in early vision system is the first step of information receiving of the vision system and the base of vision research. Based on the theory of neurophysiology, the early vision model is implemeted by the computer algorithm to simulate early vision information process of the cone pathway. The character istic of is validated, and the balance among hardware complexity, performance and precision under the condition of sufficient and essential data collection is analyzed.

关 键 词：视网膜 早期视觉 神经网络 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]