针对水下采集的图像存在色偏严重、对比度偏低的现象，本文提出一种基于自适应直方图拉伸和改进MSRCR的水下图像增强方法。该方法主要分为以下几个步骤：首先采用白平衡调整图像的颜色温度尽可能地消除色偏，其次通过设计的自适应HE分段函数对图像进行进一步的色彩平衡处理从而增强图像的视觉效果，然后使用增加了对比度增强的MSRCR对图像进行色彩恢复和对比度增强，最后通过新的AGCWD对图像进行对比度调整从而使最终获得的图像色彩平衡，明暗分明。为了证明本文方法的有效性选取了水下真实图像进行增强处理，并且选取最近几年的算法进行参考对比，实验结果证明本文算法对于水下色偏严重、对比度不高的图像具有很好的增强效果，增强后的图像质量增加、色偏减少、清晰度更高。
Aiming at the phenomenon of severe color cast and low contrast in images collected underwater, this paper proposes a method for underwater image enhancement based on adaptive histogram stretching and improved MSRCR. The method is mainly divided into the following steps: first, the white balance is used to adjust the color temperature of the image to eliminate the color cast as much as possible, secondly, the image is further color balance processing through the designed adaptive HE segmentation function to enhance the visual effect of the image, and then the MSRCR with increased contrast enhancement is used to restore the color and contrast enhancement of the image, and finally the contrast adjustment of the image is adjusted by the new AGCWD so that the final image color balance is clearly distinguished. In order to prove the effectiveness of the proposed method, the underwater real image is selected for enhancement processing, and the algorithm in recent years is selected for reference comparison, and the experimental results show that the proposed algorithm has a good enhancement effect for images with severe underwater color cast and low contrast, and the enhanced image quality is increased, the color cast is reduced and the clarity is higher.
2024,46(17): 162-167 收稿日期：2023-11-16
DOI：10.3404/j.issn.1672-7649.2024.17.029
分类号：TP391
基金项目：国家自然科学基金委员会项目（62271236）
作者简介：刘飞跃(1999-),男,硕士研究生,研究方向为机器视觉
参考文献：
[1] 李鑫杰. 水下图像清晰化增强研究[D]. 青岛：青岛大学, 2022: 6–8.
[2] ZUIDERVELD K. Contrast limited adaptive histogram equalization [C]//. Graphics Gems IV. Academic Press Professional, 1990: 474–485.
[3] IQBAL K, ODETAYO M, JAMES A, et al. Enhancing the low qualityimages using unsupervised colour correction method[C]//Pro-ceedings of 2010 IEEE International Conference on Systems Manand Cybernetics. Istanbul, Turkey: IEEE, 2010: 1703–1709.
[4] HUANG Dongmei, WANG Yan, SONG Wei. Shallow-water image en-hancement using relative global histogram stretching based on a-daptive parameter acquisition [C]//. International Conference onMultimedia Modeling, Springer, Cham, 2018.
[5] 王潇, 于炯, 杨嘉能, 等. 优化动态直方图均衡的水下图像增强方法[J]. 计算机仿真, 2022, 39(10): 259-264+439.
[6] HE K M, SUN J, TANG X O. Single image haze removal usingdark channel prior[J]. IEEE Transactions on Pattern Analysisand Machine Intelligence, 2011, 33(12): 2341-2353.
[7] CHAO L, WANG M. Removal of water scattering[C]//Proceedingsof the 2nd International Conference on Computer Engineering and Technology. Chengdu, China: 2010, (2): 35–39.
[8] FUXY, ZHUANGPX, HUANG Y, etal. A Ret-inex-based enhancing approach for single underwater image[C] //2014 IEEE International Conferenceon Image Processing(ICIP). New York: IEEE, 2014: 4572-4576.
[9] 田宁, 程莉, 元海文, 等. 基于Retinex模型的水下图像增强方法[J]. 中国科技论文, 2022, 17(11): 1281-1288.
[10] 陈浩, 刘宇涵, 冯兴华, 等. 基于色彩校正与暗通道先验算法的水下图像增强方法[J]. 西南科技大学学报, 2021, 36(3): 66-71.
[11] CHEN X Y, YU J Z, KONGSH, et al. Towardsreal-time advancement of underwater visualquality with GAN[J]. IEEE Transactions on Industrial E lectronics, 2019, 66(12): 9350-9359.
[12] Park J, Han DK, Ko H. Adaptive Weighted Multi-Discriminator CycleGAN for Underwater Image Enhancement. Journal of Marine Science and Engineering. 2019; 7(7): 200. https://doi.org/10.3390/jmse7070200.
[13] ZONG X, CHEN Z, WANG D. 2020_Local-CycleGAN: a general end-to-end network for visual enhancement in complex deep-water environment[J/OL]. Applied Intelligence, 2021: 1947–1958.
[14] LIMARE, NICOLAS, et al. "Simplest Color Balance. " Image Processing On Line, Oct. 2011, pp. 297–315. https://doi.org/10.5201/ipol. 2011. llmps-scb. Limare, Nicolas, et al. "Simplest Color Balance." Image Processing On Line, Oct. 2011, pp. 297–315.https: //doi.org/10.5201/ipol.2011.llmps-scb.
[15] LI C, TANG S, KWAN, H K , et al. Color Correction Based on CFA and Enhancement Based on Retinex With Dense Pixels for Underwater Images[J]. IEEE Access, 155732–155741.
[16] JOBSON D J, RAHMAN Z, WOODELL G A. A multiscale retinex for bridging the gap between color images and the human observation of scenes[J/OL]. IEEE Transactions on Image Processing, 1997: 965–976.
[17] HUANG, SHIH-Chia, et al. Efficient contrast enhancement using adaptive gamma correction with weighting distribution[C]//. IEEE Transactions on Image Processing, 2013; 1032–1073.
[18] MARQUES T, ALBU A. L^2UWE: A framework for the efficient enhancement of low-light underwater images using local contrast and multi-scale fusion[J]. Cornell University - arXiv, Cornell University - arXiv, 2020.
[19] KALYANA S R S, Damodaran N. Multiscale decomposition and fusion-based color contrast restoration for various water-colored environments[J]. Color ResAppl. 2021:1–28.
[20] ZHANG W, WANG Y, LI C, et al. Underwater image enhancement by attenuated color channel correction and detail preserved contrast enhancement[C]//IEEE Journal of Oceanic Engineering, 47, 718–735.
[21] ANCUTI C O, ANCUTI C, DE VLEESCHOUWER C, et al. Color balance and fusion for underwater image enhancement[J/OL]. IEEE Transactions on Image Processing, 2018: 379–393.
[22] YUAN J, CAI Z, CAO W. TEBCF: real-world underwater image texture enhancement model based on blurriness and color fusion[J/OL]. IEEE Transactions on Geoscience and Remote Sensing, 2022: 1–15.
