摘要
针对传统图像增强方法会丢失图像细节信息这一缺陷,提出了一种基于分频和奇异值分解的轮胎图像增强新方法。该方法先使用巴特沃斯低通滤波器对图像进行分频处理;然后对得到的低频和高频分量分别使用奇异值分解和线性方法进行增强;最后把处理后的低频和高频分量进行叠加得到最终增强结果。实验结果表明,新方法能在增强图像视觉效果的同时具有较好的细节保持性能。
For the existing problems of losing image details in the process of traditional image enhancement algorithms, this pa-per presented a new tire image enhancement algorithm on the basis of frequency division and singular value decomposition.Firstly, it divided the image into the low frequency part and high frequency part with Butterworth low pass filter. Then the lowfrequency part took advantage of singular value decomposition method to get enhanced, while enhanced the high frequency partby linear method. At last,it reunited the two parts which had been enhanced to get the enhanced image. The experimental re-sults demonstrate that this algorithm can enhance the visual perception effect of tire image and has a good performance of keep-ing image details.
出处
《计算机应用研究》
CSCD
北大核心
2012年第3期1178-1180,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(60803088)
陕西省自然科学基础研究计划资助项目(2009JM8018)
中央高校基本科研业务费专项资金重点资助项目(GK200901006)
陕西师范大学研究生培养创新基金资助项目(2011CXS028)
关键词
轮胎图像
奇异值
低通滤波
高斯噪声
图像增强
tire image
singular value
low pass filter
Gaussian noise
image enhancement
