基于多目标优化的图像增强算法研究  

Research on Image Enhancement Algorithm Based on Multi-Objective Optimization

作　　者：郑瑾[1] ZHENG Jin(Department of Information Engineering,Fujian Chuanzheng Communications College,Fuzhou,Fujian 350007)

机构地区：[1]福建船政交通职业学院信息工程系,福建福州350007

出　　处：《武夷学院学报》2018年第12期54-57,共4页Journal of Wuyi University

摘　　要：将图像增强问题建模为一个有约束的多目标优化问题(Multi Objective Optimization Problem, MOOP),采用多目标遗传算法求解随机共振噪声分布的最优参数。提出的算法在低质量的图像中加入适量的噪声,以提高次优增强图像的性能。性能评估部分使用实例验证了该算法的有效性。The image enhancement problem is modelled as a constrained multi-objective optimization problem(MOOP),a MOOP technology based on genetic algorithm is applied to solve the optimal solution to parameters of stochastic resonance noise distribution.In order to improve the performance of sub-optimal image enhancement,the proposed algorithm adds a suitable amount of noise to the low-quality image.Some examples are used to verify the effectiveness of the algorithm in the evaluation part.

关 键 词：图象增强 噪声 多目标优化 遗传算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]