尺度与特征引导视觉选择性注意机制模型

Model of Visual Selective Attention Mechanism Deployed by Scale and Features

针对在传统机器视觉研究中,尺度、显著性和物体识别多数被分开研究的现状,首先分析三者之间的内在联系和相互关系,得出应该在一个框架中来研究它们的结论;然后讨论视觉中的尺度空间表示方法、显著性度量方法。最后选取强度、颜色和方向三种特征以及尺度引导注意,建立一个自下而上的结合尺度与特征引导的计算模型,并给出仿真实验结果。

In traditional study of machine vision,scale,saliency and object recognition are performed respectively.Contrary to the traditional idea,they are unified in which the problems of scale in vision and the scale-space representation,the essence of the computation of saliency are discussed.Intensity,color,orientation and scale are used to construct a bottom-up computational model deployed by scale and features.And emulation experiment results are given.