| 411 | 10 | 158 |
| 下载次数 | 被引频次 | 阅读次数 |
针对地下空间低照度图像色彩偏暗、亮度低且分布不均、增强后图像色偏和噪声高等问题，研究提出了融合非物理模型的改进AM-RetinexNet图像增强算法。该算法将RGB图像转换成HSV分量，利用HSV空间相互独立性实现图像亮度增强和色彩增强处理，其中S分量利用V分量提取的相关信息进行自适应调整，V分量采用融合直方均衡化与注意力机制优化的RetinexNet进行照度分量增强处理；将处理后HSV分量转化成RGB图像，并对图像进行自适应色彩恢复，得到照度增强图像。对比实验表明，在图像的细节处理、亮度整体增强处理、图像降噪和色彩视觉修正等方面该方法表现较好，测试指标中平均互信息(MI)、标准差(STD)、结构相似性(SSIM)、平均梯度(AG)、空间频率(SF)和峰值信噪比(PSNR)最佳，均值分别可达到6.18,70.62,0.56,13.29,36.53,39.22。
Abstract：To solve the problems such as dark color, low brightness and uneven distribution, color deviation after image enhancement and high noise of low illumination image in underground space, an improved AM-RetinexNet image enhancement algorithm combining non-physical models is proposed. The algorithm converts RGB image into HSV component, and makes use of the independence of HSV space to achieve image brightness enhancement and color enhancement. Among them, S component uses the information extracted from V component for adaptive adjustment, and V component uses RetinexNet optimized by fusion of straight-square equalization and attention mechanism for illumination enhancement. Finally, the processed HSV component is transformed into RGB image, and the image is restored by adaptive color to obtain illumination enhanced image. Experimental results show that the method performs well in image detail processing, luminance enhancement, image denoising and color visual correction. The average MI, STD, SSIM, AG, SF, PSNR perform best among all metrics, with mean values respectively reaching 6.18, 70.62, 0.56, 13.29, 36.53, 39.22.
[1] 王栩，王志辉，陈昌昕，等.城市地下空间地球物理探测技术与应用[J].地球物理学进展，2021,36(5):2204-2214.
[2] LAND E H,MCCANN J J.Lightness and Retinex Theory[J].Journal of the Optical Society of America,1971,61(1):1-11.
[3] GUO C L,LI C Y,GUO J C et al.Zero-reference Deep Curve Estimation for Low-light Image Enhancement[C]//2020 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Seattle:IEEE,2020:1777-1786.
[4] KIM H,CHOI S M,KIM C S,et al.Representative Color Transform for Image Enhancement[C]//2021 IEEE/CVF International Conference on Computer Vision.Montreal:IEEE,2021:4439-4448.
[5] LI C Y,GUO J C,PORIKLI F,et al.LightenNet:A Convolutional Neural Network for Weakly Illuminated Image Enhancement[J].Pattern Recognition Letters,2018,104:15-22.
[6] 张红颖，赵晋东.HSV空间的RetinexNet低照度图像增强算法[J].激光与光电子学进展，2020,57(20):294-301.
[7] 黄辉先，陈凡浩.基于注意力机制和Retinex的低照度图像增强方法[J].激光与光电子学进展，2020,57(20):53-60.
[8] WU K,LI G J,HAN G L,et al.Color Image Detail Enhancement Based on Quaternion Guided Filter [J].The Journal of China Universities of Posts and Telecommunications,2017,24(4):40-50.
[9] 李忠海，宋笑宇，陈灿灿，等.基于Retinex算法的亮度分层图像增强算法[J].数据采集与处理，2019,34(1):41-49.
[10] BABAKHANI P,ZAREI P.Automatic Gamma Correction Based on Average of Brightness[J].Advances in Computer Science:An International Journal,2015,4(6):156-159.
[11] 罗萍，胡光宝，吕霞付，等.基于改进LOG算子的图像增强算法[J].微电子学与计算机，2019,36(9):26-29.
[12] 韩团军.改进直方图均衡化算法及FPGA实现[J].电子器件，2017,40(4):845-848.
[13] 张文霞，王春光，王海超，等.改进的CLAHE无芒隐子草叶切片图像增强[J].农机化研究，2022,44(1):25-30.
[14] 冯丽，王娟.基于Matlab的拉普拉斯图像增强算法与设计[J].电子技术与软件工程，2020(16):143-145.
[15] 刘寿鑫，龙伟，李炎炎，等.基于HSV色彩空间的低照度图像增强[J].计算机工程与设计，2021,42(9):2552-2560.
[16] 江巨浪，刘国明，朱柱，等.基于快速模糊聚类的动态多直方图均衡化算法[J].电子学报，2022,50(1):167-176.
[17] WEI C,WANG W J,YANG W H,et al.Deep Retinex Decomposition for Low-light Enhancement[J/OL].(2018-08-14) [2022-09-11].http://arxiv.org/abs/1808.04560.
[18] DAVY A,EHRET T.GPU Acceleration of NL-means,BM3D and VBM3D[J].Journal of Real-Time Image Processing,2021,18(1):57-74.
[19] ZHANG W,TANG P,ZHAO L J,et al.A Comparative Study of U-Nets with Various Convolution Components for Building Extraction[C]//2019 Joint Urban Remote Sensing Event (JURSE).Vannes:IEEE,2019:1-4.
[20] MUSTAFA N,ZHAO J P,LIU Z Y,et al.Iron ORE Region Segmentation Using High-resolution Remote Sensing Images Based on Res-U-Net[C]//2020 IEEE International Geoscience and Remote Sensing Symposium.Waikoloa:IEEE,2020:2563-2566.
[21] SHAH Z H,MüLLER M,WANG T C,et al.Deep-learning Based Denoising and Reconstruction of Super-resolution Structured Illumination Microscopy Images[J].Photonics Research,2021,9(5):B168-B181.
[22] 卢玮，高涛，王翠翠，等.Retinex理论下基于融合思想的低照度彩色图像增强算法[J].科学技术与工程，2019,19(13):151-157.
[23] 杨艳春，李娇，王阳萍.图像融合质量评价方法研究综述[J].计算机科学与探索，2018,12(7):1021-1035.
基本信息:
中图分类号:TP391.41
引用信息:
[1]王小芳,徐何方圆,刘嘉琳,等.融合非物理模型的改进AM-RetinexNet图像增强算法[J].无线电工程,2023,53(02):352-362.
基金信息:
国家自然科学基金(62006165); 成都市科技局项目(2018-YFYF-00191-SN); 四川省民办教育协会项目(MBXH21YB119); 吉利学院科研项目(2022xzky004)~~
2023-01-03
2023-01-03
2023-01-03