基于改进萤火虫优化算法的视频监控图像增强

Surveillance Video Images Enhancement Based on Improved Glowworm Swarm Optimization Algorithm

针对视频图像增强问题中连续多帧图像序列中的像素相关性,建立了一种有效的视频图像增强模型,将视频连续图像增强问题转化为从原始低质量图像像素序列到高质量增强图像像素序列的寻优问题。基本萤火虫(GSO)算法具有容易陷入极值振荡和局部最优的缺陷,为了解决这个问题,在位置更新策略中引入了全局最优个体影响因子与局部最优个体影响因子,同时为了保证迭代过程中荧光素更新的多样性,对萤火虫荧光素的挥发及增益系数进行改进,提出了改进萤火虫(IGSO)算法。结合视频图像增强问题特性,重新定义了算法的群体的输入、萤火虫的荧光素和位置更新运动方程,设定了优化目标函数准则。最后典型的道路和室内监控视频图像增强实例验证了所提出的模型和算法的可行性。

Aiming at the pixel correlation of continuous multi frame image sequences in video image enhancement,an effective video image enhancement model is established to transform the video continuous image enhancement problem into optimization problem from pixel sequence of low quality image to pixel sequence of high quality image.The basic glowworm swarm optimization(GSO)algorithm is easy to fall into the extreme value oscillation and local optimum.In order to resolve this problem,the global optimal individual impact factor and the local optimal individual impact factor are introduced in the location update strategy.At the same time,in order to ensure the diversity of fluorescein updates in the iteration process,the volatilization and gain coefficient of firefly fluorescein are improved,and an improved glowworm swarm optimization(IGSO)algorithm is proposed.Combined with the characteristics of video image enhancement,the algorithm’s swarm input,firefly’s luciferase and location update equation are redefined,and the optimization objective function criterion is set.A typical road surveillance video super-resolution reconstruction example verifies the feasibility of the proposed model and algorithm.