期刊文献+

基于稠密级联卷积神经网络的水下图像增强 被引量:4

Underwater Image Enhancement Based on Dense Cascaded Convolutional Neural Network
原文传递
导出
摘要 为了解决水下退化图像出现的色彩偏差等问题,提出一种基于稠密级联卷积神经网络的水下图像增强算法。首先将退化的水下图像从传统的红、绿、蓝(RGB)颜色空间转换到色调、饱和度、亮度(HSV)颜色空间,保持色调分量和亮度分量不变,利用级联卷积神经网络对饱和度分量增强。然后在特征提取网络编解码过程中引入了新的稠密块。稠密块将残差连接、跳跃连接和多尺度卷积结合起来,纠正颜色失真。纹理细化网络是利用了6个纹理细化单元对所得到的细化图像进一步提取特征信息。最后将通过级联卷积神经网络进行提取的S通道图与H、V通道图进行合并,得到增强的水下图像。实验结果表明,提出算法增强的水下图像的水下彩色图像质量评价平均可达到0.616875,水下图像质量测量平均可达到5.197000。对比算法表明,提出的水下图像增强算法不仅增强效果良好,且增强的结果更符合人类视觉习惯。 To solve the low contrast problem of underwater degraded images,an underwater image enhancement algorithm based on a deep cascaded convolutional neural network is proposed.First,the degraded underwater image is converted from traditional red,green,and blue to hue,saturation,and value color space,which retains the hue and lightness component without changes,and the cascaded convolutional neural network is employed to examine the saturation component improvement.New dense blocks are introduced in the process of feature extraction network encoding and decoding.The dense block combines residual connection,skip connection,and multiscale convolution to correct color distortion.The texture refinement network employs six texture refinement units to extract feature information from the refined image.Finally,the S-channel image is extracted using the cascaded convolutional neural network,which is combined with the H-and V-channel images to achieve an improved underwater image.The experimental findings reveal that the average underwater color image quality estimation of underwater images improved using the proposed algorithm can reach 0.616875,and the average underwater image quality measurement can reach 5.197000.The comparison algorithm findings reveal that the proposed underwater image enhancement algorithm not only has a good improvement effect but also ensures the improved images are in line with human vision.
作者 陈清江 解亚丽 Chen Qingjiang;Xie Yali(School of Science,Xi’an University of Architecture and Technology,Xi’an 710055,Shaanxi,China)
出处 《激光与光电子学进展》 CSCD 北大核心 2022年第22期227-236,共10页 Laser & Optoelectronics Progress
基金 国家自然科学基金(61403298) 陕西省自然科学基金(2015JM1024) 陕西省教育厅专项科研计划项目(2013JK0586)。
关键词 机器视觉 水下图像 卷积神经网络 编码解码框架 计算机视觉 稠密块 maching vision underwater image convolutional neural network coding and decoding framework computer vision dense block
  • 相关文献

参考文献8

二级参考文献41

共引文献58

同被引文献33

引证文献4

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部