用二元小波滤波器作图像相关识别

Pattern Correlation Recognition Using a Binary Wavelet Filter

本文提出用二元小波滤波器作图像相关识别的技术。利用小波变换对图像处理的优越性 ,对小波滤波函数进行二元化处理 ,成为二元带通滤波器 ,并应用于旋转图像的识别。本文进行了理论阐述和计算机模拟实验 ,结果证实了设计的可行性 。

Pattern correlation recognition using a binary wavelet filter (BWF) is proposed. By making use of the advantages of wavelet transform for pattern processing, the function of wavelet filter is binarized. The wavelet filter becomes to be a binarized bandpass filter, and is used to recognize rotational patterns. Theoretical exposition and computer simulation are provided. The results have verified the correctness of the design, and yield a better correlation output than the classical joint transform correlation.