摘要
图像边缘细节富含高频成分,鉴于小波包分析具有对高频分辨率高的特点,提出一种应用小波包变换的图像加权广义模糊增强算法。对图像进行小波包分解,分别对各分解子图像进行基于线性广义模糊算子(LGFO)的模糊增强处理,重构前对各增强子图像赋予不同的权值,通过小波包图像重构实现对原始图像的加权模糊增强处理,获得细节丰富、对比度强的高质量增强图像。在基于边缘测度与噪声标准差的增强图像质量评估标准下,算法实现了模糊参数的自适应寻优。实验表明,增强图像轮廓准确、细节丰富。
Edge details of an image are rich in high frequency components. In view of the character of high-frequency high resolution for wavelet packet analysis, an algorithm is put forward for image weighted generalized fuzzy enhancement using wavelet packet transform. The selected image is decomposed by wavelet packet transform. Each decomposed sub-image is enhanced using fuzzy enhancement based on linear generalized fuzzy operator (LGFO), respectively. Before image reconstruction, different weight coefficients are assigned to enhanced sub-images. The process of weighted fuzzy enhancement for the original image is realized by wavelet packet reconstruction. Then the high-quality enhanced image with rich details and strong contrast can be obtained. The algorithm can achieve adaptive optimization for the fuzzy parameter using the quotient between edge measure degree and noise standard deviation as the quality standard of enhanced image. The experiments show that the edge contour of enhanced image is accurate, rich-details.
出处
《燕山大学学报》
CAS
2009年第6期478-483,共6页
Journal of Yanshan University
基金
教育部博士点基金资助项目(2006021600)
关键词
图像广义模糊增强
小波包变换
线性广义模糊算子
边缘测度:噪声标准差
image generalized fuzzy enhancement
wavelet packet transform
linear generalized fuzzy operator (LGFO)
edge measure degree
noise standard deviation