摘要
根据电视导引头图像处理系统的要求,基于小波变换的多分辨分析思想,提出一种多尺度小波变换的阈值化去噪(图像恢复)算法并与传统去噪算法进行了比较。该算法在门限化小波细节系数中采用平滑过渡区的思想,克服了传统阈值化处理的小波系数不连续性以及存在恒定的偏差的缺点。实验证明该算法在图像去噪方面是可行的和有效的,满足实时处理的要求,而且能够方便地产生各种分辨率的图像。
A thresholding de-noise algorithm is presented based on multilevel wavelet analysis according to the requirement of TV seeker image process system, and its de-noise is a very effective than those presented by traditional techniques. This algorithm overcomes the disadvantages in traditional wavelet transform thresholding algorithm whose wavelet transform coefficients is discontinuous and exists invariable error through using interpolation in thresholding wavelet detail coefficients. It can be seen from results of simulation that the algorithm is a very effective in de-noise, and the algorithm is a practical and promising image de-noised algorithm.
出处
《弹箭与制导学报》
CSCD
北大核心
2004年第1期13-15,共3页
Journal of Projectiles,Rockets,Missiles and Guidance
