应用综合鉴别函数实现畸变不变图像识别

Distortion Invariant Pattern Recognition in Terms of Synthetic Discriminant Function

摘要: 针对传统图像识别方法对目标图像的尺寸、方位敏感问题,采用综合鉴别函数(SDF)实现比例和旋转不变图像识别.给出了应用综合鉴别函数实现畸变不变图像识别的基本原理,研究了加权系数的计算,提出了用图像序列集tn(x,y)合成综合鉴别函数的方法.通过计算机模拟和光学实验,证明了该方法的可行性.

Abstract: Classic pattern correlators are sensitive to scale variation and rotation of the target images.Synthetic discriminant function(SDF) are used instead for distortion invariant pattern recognition.Mathematical principle of SDF and method are provided to its coefficient of training images tn(x,y).Computer simulation and experimental results showed that the SDF is feasible for distortion-invariant pattern recognition.