小波融合技术在运动模糊图像复原中的应用  

Wavelet Fusion Technology Used in Application of Motion Blurred Image Restoration

作　　者：赵环旭 滕青芳[1] 

机构地区：[1]兰州交通大学自动化与电气工程学院,甘肃兰州730070

出　　处：《电视技术》2014年第5期9-12,共4页Video Engineering

摘　　要：由于物体和成像装置之间会产生相对位移,通常使获取的图像变得模糊不清。提出将改进后的小波融合技术运用在运动模糊图像复原中,运用该思想来进行图像复原。即利用Radon变换来获得模糊角度,运用微分自相关法估计出模糊尺度,接着选取两种经典复原方法进行图像复原,最后用小波融合技术对复原图像进行融合实验。从实验结果可以看出,相较于经典复原方法,该方法得到的客观评价更好,并且能保留一定的融合图像的边缘细节,具有实际参考价值。The obtained image is always blurred ,because of the relative displacement between the object and the imaging device. Improved wavelet fu-sion technique is proposed to use in the motion blurred image restoration. The method is applied for image restoration. The blur direction of image is ob-tained by using Radon transform,and the blur length of image is estimated by using the differential autocorrelation method. After that ,two classical resto-ration methods are selected for image restoration. Finally ,wavelet fusion technique is used to recover image fusion experiment. Recovery can be seen from the experimental results ,compared with classical methods, this method can get better objective evaluation, and can retain a certain fusion of the edge de- tails of image. This method has a certain practical reference value.

关 键 词：RADON变换 微分自相关 小波融合技术 

分 类 号：TN911.73[电子电信—通信与信息系统] N391.4[电子电信—信息与通信工程]