一种改进的结合边缘检测的去雾新方法

摘要/Abstract

摘要：

摘要：暗原色去雾算法复原的图像会出现边缘效应，为此提出了一种改进的基于暗原色理论的去雾算法.首先计算出图像的暗色图，并使用小波分解算法将暗色图像向下采样一次；然后在小波分解的平滑部分计算大气光和图像的暗原色，提取出平滑部分的边缘信息，并使用形态学对其进行膨胀获得相应的二值边缘图像；接着对二值边缘图像所在区域的暗原色及非边缘处的暗原色分别进行优化；最后，将优化后的暗原色使用小波重构算法向上采样，并据此计算出图像透射率，结合去雾模型复原到清晰的无雾图像.实验证明，该方法在很好地去除边缘效应的同时，极大地减少了算法时间复杂度，满足实时要求.

关键词: 图像去雾算法, 暗原色优先理论, 小波, 边缘提取

Abstract:

Abstract： The images restored by using the dark channel prior theory have haze edges. Therefore, a novel improved image clearness method based on dark channel prior was proposed to solve this insufficient. First, dark color image was obtained and decomposed by wavelet. Next, the smooth portion from wavelet decomposition was used as the target image, in which global atmospheric optical A and the dark channel value were calculated, and then, the edge information of target image was extracted and dilated. After that, according to the expanded binary edge image, the dark channel value in edges and that in smooth were optimized, respectively. Finally, according to the transmission map calculated using the optimized dark channel values, the fogfree image was obtained by using the physical model. The experiment results show that this method can effectively restore the fogging images and greatly reduce the time complexity brought by the soft matting algorithm.

Key words: clearness algorithm, dark channel prior, wavelet, edge detection

中图分类号:

TP 391

黄大荣,方周，赵玲. 一种改进的结合边缘检测的去雾新方法[J]. 上海交通大学学报（自然版）, 2015, 49(06): 861-867.

HUANG Darong,FANG Zhou，ZHAO Ling. An Improved Defogging Algorithm Combined with Edge Detection[J]. Journal of Shanghai Jiaotong University, 2015, 49(06): 861-867.

参考文献

［1］DING Meng, TONG Ruofeng. Efficient dark channel based image dehazing using quadtrees［J］. Science China Information Sciences, 2013, 9(56):19.

［2］Tarel J P,Nicolas H. Fast visibility restoration from a single color or gray level image［C］∥ICCV. Kyoto, Japan: IEEE, 2009:22012208.

［3］NARASIMHAN S G, NAYAR S K.Interactive (de) weathering of an image using physical models ［C］∥Proceedings of the ICCV Workshop on Color and Photometric Methods in Computer Vision. Nice, France: IEEE, 2003: 13871394.

［4］Narasimhan S G, Nayar S K.Contrast restoration of weather degraded images［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2003, 25(6):713724.

［5］He Kaiming, Sun Jian.Single image haze removal using dark channel prior［C］∥CVPR. Washington DC: IEEE, 2009:19561963.

［6］Xie Bin, Guo Fan. Improved single image dehazing using dark channel prior and multiscale retinex ［C］∥Proceedings of IEEE Conference on 2010 International Conference on Intelligent System Design and Engineering Application. Changsha, China: IEEE, 2010: 848851.

［7］Yang Hongbo, Wang Jiuhe. Color image contrast enhancement by cooccurrence histogram equalization and dark channel prior ［C］∥Proceedings of IEEE Conference on 2010 3rd International Congress on Image and Signal Processing. Yantai, China: IEEE, 2010: 848851.

［8］He Kaiming, Sun Jian. Guided image filtering［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(6): 13971409.

［9］Tan R. Visibility in bad weather from a single image［C］∥CVPR. Anchorage, AK: IEEE, 2008:18.

[1]	白子轩1, 马艺馨1,2, 孔志斌3, 薛珊4. 基于胸阻抗测量的肺灌注/心脏运动和肺通气同步监测可行性实验研究[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(1): 81-90.
[2]	BALASUBRAMANIAN S1, NARUK Mahaveer Singh2, TEWARI Gaurav3. 基于经验小波变换优化自适应混合滤波器的心电信号去噪[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(1): 66-80.
[3]	朱昶胜，朱丽娜. 基于经验小波变换、循环神经网络和误差校正的短期风速预测[J]. J Shanghai Jiaotong Univ Sci, 2024, 29(2): 297-308.
[4]	段继忠, 钱青青. 基于SIDWT和迭代自一致性的快速并行成像重建方法[J]. 上海交通大学学报, 2023, 57(5): 582-592.
[5]	王火平, 王德洋, 彭怡锦, 王巍巍. 振动监测在导管架平台结构安全保障方面的几种应用[J]. 海洋工程装备与技术, 2023, 10(4): 78-83.
[6]	郝国成, 张必超, 锅娟, 张雅冰, 石光耀, 王盼盼, 张薇. 高质量LMSCT时频分析算法及其在雷达信号目标检测中的应用[J]. 上海交通大学学报, 2022, 56(2): 231-241.
[7]	李春祥，裴杨从琪，殷潇. 基于Hermite组合核EMD-WT-LSSVM的非平稳非高斯风压预测[J]. 上海交通大学学报, 2019, 53(10): 1249-1258.
[8]	李春祥,殷潇. 基于小波支持向量机的非高斯空间风压内外插预测[J]. 上海交通大学学报（自然版）, 2018, 52(11): 1516-1523.
[9]	高筱婷，杨东升. 基于数据驱动的智能电器运行状态监测方法 [J]. 上海交通大学学报（自然版）, 2017, 51(9): 1104-1110.
[10]	王向红1a,1b，尹东1a，胡宏伟1a，毛汉领2. 基于小波包和独立分量分析的微弱多源故障声发射信号分离[J]. 上海交通大学学报（自然版）, 2016, 50(05): 757-763.
[11]	李庆1,LIANG Steven Y1,2,杨建国1. 谐波小波样本熵与HMM模型的轴承故障模式识别[J]. 上海交通大学学报（自然版）, 2016, 50(05): 723-729.
[12]	薛海建, 郭晓松, 张东方, 周召发. 基于四元数的捷联惯导惯性系晃动基座自对准算法[J]. 上海交通大学学报, 2016, 50(03): 419-424.
[13]	纪良波. 基于小波神经网络熔丝堆积三维打印精度预测模型[J]. 上海交通大学学报（自然版）, 2015, 49(03): 375-378.
[14]	田福庆1，罗荣1，李万2，谢勇1. 改进的谐波小波包峭度图及其应用[J]. 上海交通大学学报（自然版）, 2014, 48(1): 39-44.
[15]	杨光，吴静怡. 基于小波变换和多元回归的航天器热循环试验系统温度均匀性分析[J]. 上海交通大学学报（自然版）, 2014, 48(09): 1346-1350.