摘要
小波理论在图像去噪领域得到了广泛应用,在研究传统小波软阈值算法的基础上,分析了该算法的优点与不足之处,针对传统算法易造成背景信号模糊以及低频信号的损失的缺陷,在阈值选取时,提出了分层调整因子的概念,经过严格的理论证明,当调整因子取值1/2时,对不同的图像处理有较好的稳定性,且算法实现与小波基选取无关。研究结果表明:新算法在一定程度上弥补了传统算法的缺陷,去噪图像的频谱比传统算法宽广,低频信号保留较好,图像背景清晰平滑,去噪的视觉效果明显,而且具有较好的可操作性,易于实现。其应用于临场感系统取得了较好的效果,显示出新算法具有良好的应用前景。
Wavelet has been applied widely on the field of image de-noising. Traditional soft- threshold algorithm of de-noising was researched, and the advantages and disadvantages of traditional algorithm were analyzed. Then in view of the disadvantages, bringing forward the conception of adjustment factor, through proved precisely, when the factor was 1/2, to many kinds of different images, the result of processing was stable, and the algorithm was unrelated to the wavelet bases. Results of the theory and experiment showed that new algorithm had many advantages, such as operable, implement easily and great ability of de-noising in low frequency. So traditional algorithm was remedied in a certain extent. Effect of the new algorithm is better than the traditional algorithm in telepresence system.
出处
《计算机工程与设计》
CSCD
北大核心
2006年第2期350-351,355,共3页
Computer Engineering and Design
关键词
图像去噪
小波
阈值
软阈值算法
临场感
image de-noising
wavelet
threshold
soft-threshold algorithm
telepresence
