LI Mingdong, ZHANG Juan, WU Shiqian, et al. License plate image deblurring algorithm based on RANSAC transform[J]. 2020, (2): 153-156+160. DOI: 10.13873/j.1000-9787(2020)02-0153-04.
基于RANSAC变换的车牌图像去模糊算法
摘要
针对监控视频中运动模糊车牌图像的复原问题
提出一种基于随机抽样一致性(RANSAC)的车牌图像去模糊算法。首先
在倒频谱域对模糊图像实施Canny边缘检测
使用RANSAC变换来估计模糊方向;然后
通过模糊图像倒频谱域的零点分量来估计模糊长度;最后
结合Richardson-Lucy(RL)算法和L0正则化先验各自的优点
对模糊车牌图像进行复原。实验结果表明:提出的算法不论主观还是客观上
在运动模糊车牌图像上达到了很好的复原效果
复原图像得到的PSNR和SSIM值都有了极大地提高
并且耗时少
效率高。
Abstract
Aiming at the restoration problem of motion blurred license plate images in surveillance video
a license plate image deblurring algorithm based on random sampling consistency( RANSAC) is proposed. Firstly
Canny edge detection is performed on the blurred image the cepstrum domain
and RANSAC transform is used to estimate the blur direction. Then
the blur length is estimated by the zero component of the blur image in cepstrum domain.Finally
the blurred license plate image is restored with the advantages of Richardson-Lucy( RL) algorithm and L0 regularization prior. The experimental results show that the proposed algorithm achieves a good recovery effect on the motion blurred license plate images both subjectively and objectively. The PSNR and SSIM values obtained from restored images is greatly improved