基于小波变换和方向微分的运动模糊方向鉴别研究  

Motion Blurred Direction Identification Based on Wavelet Transform and Directional Derivation

作　　者：张爱萍[1] 李浪 韩亮 ZHANG Aiping;LI Lang;HAN Liang(Xi'an University of Architecture and Technology,Xi'an Shaanxi 710055)

机构地区：[1]西安建筑科技大学理学院,陕西西安710055

出　　处：《软件》2021年第1期59-63,共5页Software

基　　金：陕西省碑林区科技局项目资助(GX1902)。

摘　　要：针对最小方向微分算法(MinimumDirectionalDerivation,MDD)在运动模糊方向鉴别中计算量大的问题,提出利用小波变换和方向微分相结合的算法鉴别运动模糊方向。首先引入小波变换获得运动模糊图像的低频部分,缩小了计算范围,减小了计算量;然后在利用基于方向微分的加权平均法鉴别模糊方向时,先取步长为10°粗略得到模糊方向,在此模糊方向±10°范围内以0.5°为步长精确鉴别出模糊方向。该方法既能减小计算量,又能提高鉴别精度。仿真结果表明本文算法鉴别运动模糊方向误差小,实时性好,应用范围广。Concerning large amount of computations by the minimum directional derivation(MDD)algorithm,a method in combination with wavelet transform and directional derivation is put forward to identify the blur direction from motion blurred image.Firstly,wavelet transform is introduced to obtain the low-frequency component of the motion blurred image,which contains the main information of the image and reduces the computational efforts compared with the original motion blur image.Secondly,to identify the blur direction based on the directional derivation and weighted average method,which the rough blur direction is obtained according to the step length of 10°,and then the accurate blur direction is obtained in the range of±10°by taking 0.5°as the step length.The proposed scheme can not only reduce the computational efforts,but also improve the identification accuracy.Simulation results show that this proposed algorithm has small error of motion fuzzy direction,good real-time performance and wide application range.

关 键 词：小波变换 方向微分 加权平均 步长 运动模糊方向 图像复原 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]