基于圆对称性的离焦模糊图像还原方法  

Restoration of Defocus Blurred Image Based on Circular Symmetry

作　　者：马云燕[1] 马燕[1] 李顺宝 张相芬[1] 

机构地区：[1]上海师范大学信息与机电工程学院,上海200030 [2]上海师范大学数理学院,上海200030

出　　处：《电视技术》2014年第23期9-12,共4页Video Engineering

基　　金：国家自然科学基金项目(61373004)

摘　　要：针对离焦模糊图像,圆盘离焦模型和高斯离焦模型的还原效果并不理想,提出基于圆对称性的离焦模糊图像还原算法。通过刃边函数估算模糊半径,利用信噪比还原图像,进而计算图像的信号功率谱。提出改进的DCT分块加权清晰度评价函数,对DCT系数作指数加权优化,并根据圆对称性模型检测出的多个模糊半径还原出多个图像,并筛选出最清晰图像作为还原结果。实验结果显示,与传统的还原方法相比,该方法提高了还原图像的清晰度和分辨率。For the defocus blurred image ,the restoration result of disc defocus model and Gaussian defocus model is unsatisfied. This paper presents defocus blur image restoration algorithms based on circular symmetry. Blur radius is estimated by blade edge function. The image is restored by using SNR, and then the signal power spectrum of the image is calculated. It presents clarity evaluation function based on improved weighted DCT block. The DCT coefficients are calculated by exponential weighting. The clearest image is chosen as a result by filtering several images which are restored by blur radius based on circular symmetry model. The experimental results show that, compared with the traditional restoration method, the proposed method improves the image clarity and resolution.

关 键 词：离焦模糊图像 还原 清晰度评价方法 权重 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]