摘要
为了提高变分Retinex理论模型的增强效果和计算速度,提出一个用于彩色图像增强的模型.该模型采用全变分理论,使用L1范数替代原有变分模型中对光照图像进行约束的L2范数;使用切分Bregman迭代算法进行求解,由于L1范数与L2范数的凸性,该算法可以获得全局最优解.与原有变分模型进行比较的实验结果表明,文中模型在增强效果基本相近的情况下,计算速度提高了近40倍.
In this paper,a fast algorithm based on Retinex theory for color image enhancement is proposed.A total variational framework is employed,in which L1 norm instead of the original L2 norm is adopted.As a convex functional,the proposed model gives the solution of global optimization.The split Bregman method is employed to compute our model with high efficiency.In addition to its computational efficiency,our model is also good at preserving texture details and color information of the image.The proposed method is compared with two popular methods in terms of enhancement performance and computational efficiency.The experimental results show that the proposed method is nearly 40 times faster than the original methods with comparable enhancement performance.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2010年第10期1777-1782,共6页
Journal of Computer-Aided Design & Computer Graphics
基金
国家国际科技合作项目(2009DFA12290)
