Journal of Computer Applications ›› 2019, Vol. 39 ›› Issue (4): 1162-1169.DOI: 10.11772/j.issn.1001-9081.2018091979

Previous Articles     Next Articles

Weakly illuminated image enhancement algorithm based on convolutional neural network

CHENG Yu, DENG Dexiang, YAN Jia, FAN Ci'en   

  1. Electronic Information School, Wuhan University, Wuhan Hubei 430072, China
  • Received:2018-09-26 Revised:2018-10-26 Online:2019-04-10 Published:2019-04-10
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (61701351).

基于卷积神经网络的弱光照图像增强算法

程宇, 邓德祥, 颜佳, 范赐恩   

  1. 武汉大学 电子信息学院, 武汉 430072
  • 通讯作者: 邓德祥
  • 作者简介:程宇(1996-),男,江西景德镇人,硕士研究生,主要研究方向:图像去雾、弱光照图像增强;邓德祥(1961-),男,湖北荆州人,教授,博士,主要研究方向:图像识别、目标跟踪、智能物联网;颜佳(1983-),男,湖北天门人,副研究员,博士,主要研究方向:图像质量评价、图像美学评价、目标跟踪;范赐恩(1975-),女,浙江慈溪人,副教授,博士,主要研究方向:图像超分辨率重建、图像质量评价。
  • 基金资助:
    国家自然科学基金资助项目(61701351)。

Abstract: Existing weakly illuminated image enhancement algorithms are strongly dependent on Retinex model and require manual adjustment of parameters. To solve those problems, an algorithm based on Convolutional Neural Network (CNN) was proposed to enhance weakly illuminated image. Firstly, four image enhancement techniques were used to process weakly illuminated image to obtain four derivative images, including contrast limited adaptive histogram equalization derivative image, Gamma correction derivative image, logarithmic correction derivative image and bright channel enhancement derivative image. Then, the weakly illuminated image and its four derivative images were input into CNN. Finally, the enhanced image was output after activation by CNN. The proposed algorithm can directly map the weakly illuminated image to the normal illuminated image in end-to-end way without estimating the illumination map or reflection map according to Retinex model nor adjusting any parameters. The proposed algorithm was compared with Naturalness Preserved Enhancement Algorithm for non-uniform illumination images (NPEA), Low-light image enhancement via Illumination Map Estimation (LIME), LightenNet (LNET), etc. In the experiment on synthetic weakly illuminated images, the average Mean Square Error (MSE), Peak Signal-to-Noise Ratio (PSNR) and Structural SIMilarity index (SSIM) metrics of the proposed algorithm are superior to comparison algorithms. In the real weakly illuminated images experiment, the average Natural Image Quality Evaluator (NIQE) and entropy metric of the proposed algorithm are the best of all comparison algorithms, and the average contrast gain metric ranks the second among all algorithms. Experimental results show that compared with comparison algorithms, the proposed algorithm has better robustness, and the details of the images enhanced by the proposed algorithm are richer, the contrast is higher, and the visual effect and image quality are better.

Key words: weakly illuminated image enhancement, Retinex model, derivative image, Convolutional Neural Network (CNN), blind image quality assessment

摘要: 针对现有的弱光照图像增强算法强烈依赖于Retinex理论、需人工调整参数等问题,提出一种基于卷积神经网络(CNN)的弱光照图像增强算法。首先,利用四种图像增强手段处理弱光照图像得到四张派生图,分别为:限制对比度自适应直方图均衡派生图、伽马变换派生图、对数变换派生图、亮通道增强派生图;然后,将弱光照图像及其四张派生图输入到CNN中;最后经过CNN的激活,输出增强图像。所提算法直接端到端地实现弱光照图像到正常光照图像的映射,不需要按照Retinex模型先估计光照图像或反射率图像,也无需调整任何参数。所提算法与NPEA(Naturalness Preserved Enhancement Algorithm for non-uniform illumination images)、LIME(Low-light image enhancement via Illumination Map Estimation)、LNET(LightenNet)等算法进行了对比。在合成弱光照图像的实验中,所提算法的均方误差(MSE)、峰值信噪比(PSNR)、结构相似度(SSIM)指标均优于对比算法。在真实弱光照图像实验中,所提算法的平均自然图像质量评价度量(NIQE)、熵指标为所有对比方法中最优,平均对比度增益指标在所有方法中排名第二。实验结果表明:相对于对比算法,所提算法的鲁棒性较好;经所提算法增强后,图像的细节更丰富,对比度更高,拥有更好的视觉效果和图像质量。

关键词: 弱光照图像增强, Retinex模型, 派生图, 卷积神经网络, 自然图像质量评价

CLC Number: