摘要
本文介绍了一种新的旋转不变性目标识别方法。不变性特征是图像子波(wavelet)变换所得的模值,由于子波变换提取了图像中不同区域的不变性待征,使得各样本的特征矢量间的欧氏距离在绝大多数情况下比Zernike矩得到的要大得多,旋转不变性目标识别由Hamming网络实现,实验中,使用子波变换对26个英文字母相应的624种不同方向的字母进行了识别,识别的成功率为99.7%,而使用Zernike矩,识别的成功率为98%。
A new set of rotation invariant features for image recognition is introduced. The features are the magnitudes of complex Wavelet Transform (WT) of the image. The proposed method is more advantageous than Zernike Moment (ZM), for example, the Hamming Distance (HD) between the feature vectors of the different classes are larger because WT can extract the corresponding local features in the different areas. The performance of the method is expermentally tested on a 26-class data set involving differently oriented binary images. The set consists of 624 images of all English characters. Using Hamming network 99.7% and 98% classification accuracies are obtained respectively by WT and ZM.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
1995年第1期14-20,共7页
Pattern Recognition and Artificial Intelligence
基金
国家攀登计划认知科学(神经网络)重大关键项目
关键词
目标识别
子波变换
模式识别
Rotation Invariant Pattern Recognition, Wavelet, Transform, Zernike Moment, Hamming Network, Pattern Recognition.
