基于SCBSO算法的低照度纹理图像增强方法

Low-Illuminance Texture Image Enhancement Method Based on SCBSO Algorithm

针对纹理图像处理过程中,采集图像包含大量噪声而影响处理结果的问题,以及天牛须群优化(BSO)算法易陷入局部优解的问题,提出一种基于正余弦策略的改进天牛须群优化(SCBSO)算法,并将该算法应用在低照度纹理图像增强中。首先引入logistic模型增加初始解群的多样性;其次结合正余弦策略对BSO算法的搜索策略进行改进,加入时变加速因子实现参数自动更新,提升BSO算法的收敛速度和搜索精度;最后利用SCBSO算法结合染色体结构实现对图像最优灰度分布的精确搜索。在标准函数的测试中,SCBSO算法在两种类别函数下的运行时间较原算法缩短了16.56%和14.78%,增强后图像的对比度更强,自然特性保存得更好。SCBSO算法与对比算法相比,明度顺序误差(LOE)降低了37.8%,视觉信息保真度增长了15.3%,PSNR提高了12.9%,在去噪的同时很好地保留图像的纹理特征。

To overcome the problems that captured image contains a considerable noise and affects the processing result,and beetle swarm optimization(BSO)algorithm is easy to fall into the local optimal solution in texture image processing,an improved sine cosine strategy based beetle swarm optimization(SCBSO)algorithm is proposed and applied to low-illuminance texture image enhancement.First,a logistic model is introduced to increase the diversity of the initial solution group.Then,combined with the SCBSO,the search strategy of the algorithm is improved and time-varying acceleration factor is added to realize the automatic updating of the parameters,thereby improving the convergence speed and search accuracy.Finally,the improved SCBSO algorithm is combined with the chromosome structure to achieve an accurate search for the optimal grayscale distribution of the image.In a standard function test,the SCBSO algorithm shortens the performance time by 16.56%and 14.78%compared to the original algorithm under two categories of functions.The image contrast is enhanced and the natural characteristics are better preserved.As compared with the comparison algorithm,the lightness order error(LOE)of the SCBSO algorithm is reduced by 37.8%,the visual information fidelity(VIF)is increased by 15.3%,and the peak signalto-noise ratio(PSNR)is increased by 12.9%.The textural features of the image are well preserved during denoising.