基于突变粒子群算法的图像自适应增强

Image Adaptive Enhancement Based on Mutational Particle Swarm Optimization

利用突变粒子群算法自动获取图像非线性增强函数的最佳变换参数,达到图像增强的效果。该算法基于粒子群算法原理,采用针对图像质量评价效果的新适应度函数(包括方差、信息熵、紧致度、信噪改变量以及像素差别五要素),提出一种基于突变机制的粒子群算法,有效增大粒子间的差异性和非均匀性,打破平衡态,从而增强系统内动力以提高系统进化的效率。实验表明,该算法具有较高的自适应性,即避免了陷入局部极小,加快了收敛速度,且增强质量评价明显提高。

Automatically obtain the best transformation parameters of image nonlinear enhancement function based on the mutational particle swarm algorithm is proposed. This algorithm is based on particle swarm optimiza- tion principle and using a new fitness function suit for image quality evaluation （ including the variance, information entropy, the firmness, the changing of signal and noise and the pixel difference） , increasing the difference and non-uniformity between the particles effectively, and breaking the equilibrium, thereby enhancing the power system even to improve the efficiency of the system evolution. The experiments show that the algorithm has a higher self- adaptive, convergence speed up, and enhance the quality assessment significantly improved.