摘要
本文介绍了一种基于曲波变换和遗传算法的图像去噪方法,该方法利用软阈值规则调节噪声图像的曲波系数,以达到去除图像噪声的目的,去噪时使用遗传算法和广义交叉验证准则搜索最优的阈值。基于Lena和Barbara图像的实验结果表明,与小波图像去噪相比,曲波去噪后图像峰值信噪比(PSNR)和视觉效果有较大提高,特别是对图像边缘的恢复上效果明显。
In this paper, an image denoising method is presented based on curvelet transform and genetic algorithms （GAs）. In such a method, the soft thresholding algorithms are employed to threshold the eurvelet coefficients of the noise images; genetic algorithms and general cross validation （GCV） are used to search the optimal threshold. Experiment results performed on Lena and Barbara images show that the PSNR and vision of the curvelet denoised images improve a lot, especially at the edge of the images, as compared to the wavelet method.
出处
《中山大学研究生学刊（自然科学与医学版）》
2009年第2期105-114,共10页
Journal of the Graduates Sun YAT-SEN University(Natural Sciences.Medicine)
关键词
曲波
脊波
小波
图像去噪
遗传算法
广义交叉准则
Curvelet
Ridgelet
Wavelet
Image Denoising
Genetic Algorithms （ GAs ）
General Cross Validation （GCV）
