中国激光, 2012, 39 (4): 0409002, 网络出版: 2012-03-17
基于相关输出相似性度量的目标识别算法
Algorithm of Object Recognition Based on Similarity Measurement of Correlation Output
图像处理 联合变换相关器 相似性度量 相关峰 image processing joint transform correlator similarity measurement correlative peak
摘要
为了提高光学联合变换相关器(JTC)的目标识别能力,基于相似性度量原理,提出了一种新颖的相关输出图像数字处理算法。实验结果表明,该算法能够准确地识别旋转畸变-4°到+10°及缩放比例畸变70%到140%范围内的待识别目标,极大地提高了光学联合变换相关器的畸变不变容忍度。
Abstract
In order to improving the ability of objective recognition with optical joint transform correlator (JTC). A novel algorithm for the correlator output image is proposed. The experimental results show that this algorithm can accurately identify the object with rotated range form -4° to +10° and scale range from 70% to 140%, improve tremendously the distortion unchanged tolerance.
胡茂海. 基于相关输出相似性度量的目标识别算法[J]. 中国激光, 2012, 39(4): 0409002. Hu Maohai. Algorithm of Object Recognition Based on Similarity Measurement of Correlation Output[J]. Chinese Journal of Lasers, 2012, 39(4): 0409002.