摘要
提出了一种基于小波包分解的提取图像特征的算法,将图像在不同尺度下的小波包变换模极大值组成一个矩阵,采用奇异值分解得到该矩阵的奇异值矢量作为描述信号调制样式的特征向量。该算法克服了传统的通过各种频域变换提取图像特征时对图像特征难以充分描述的不足,提取的特征向量的维数相对不高,便于实现。计算机仿真结果表明,这种方法具有稳健的抗噪性和良好的扩展性。
This paper presents an algorithm to extract the image characteristics based on the wavelet packet decomposition, that uses the singular value vector to describe different modulation style signals, which can be gained by decomposing the modulus maxima of signal＇s wavelet packet transform. This algotrithm overcomes the shortcoming that is difficult to describe image characteristics by frequency domain transform feature while extracting the image characteristics, the extracted characteristics vector has low dimension, and is easy to be realized. The result of computer simulation indicated that the method is robust to anti-noise and is easy to expand.
出处
《舰船电子对抗》
2007年第4期92-95,104,共5页
Shipboard Electronic Countermeasure
关键词
小波包分解
图像消噪
图像压缩
图像融合
wavelet packet decomposition
image denoising
image compressing
image fusion
