摘要
为了有效去除水下图像倒影,而不影响水下图像细节信息完备性,研究一种基于平滑化区域填充的水下图像倒影去除方法。通过Lab颜色模型和K-means聚类识别水下图像倒影区域,计算含倒影水下图像各个像素RGB颜色分量的最小值,并导进和含倒影水下图像的灰度图中,实现图像增强处理,在需填充区域中心设置一个符合平滑填充条件的窗口尺寸,基于平滑化区域填充去除倒影,保证倒影区域去除后,图像细节信息不出现严重缺失。研究结果显示,某水下图像倒影区域去除后,水下图像有价值细节与原图有价值细节相似度较高,信息量完备性显著,不存在颜色失真情况,图像细节信息较完备。
The completeness of underwater image detail information is facile to affect by underwater image reflection. Therefore, an underwater image reflection removal method based on smoothing region filling was studied in this paper. According to lab color model and K-means clustering, the reflection region of underwater image was recognized. The minimum RGB color component of each pixel of the underwater image(including the underwater reflection) was calculated. Meanwhile, the calculation results were imported into the gray image of the underwater image with the reflection, thus enhancing the image. In the center of the area to be filled, a window size was set which meets the smooth filling conditions. By smoothing the region, the removed reflection was filled for the integrity of image detail information. The results show that this method has significant completeness of detail information.
作者
吴函
唐磊生
夏昕阳
徐红丽
WU Han;TANG Lei-sheng;XIA Xin-yang;XU Hong-li(Shenyang Institute of Automation,Chinese Academy of Sciences,Shenyang Liaoning 110016,China;Institutes for Robotics and Intelligent Manufacturing,Chinese Academy of Sciences,Shenyang Liaoning 110169,China;University of Chinese Academy of Sciences,Beijing 100049,China;Faculty of Robot Science and Engineering,Northeastern University,Shenyang Liaoning 110819,China;Information&Control Engineering Faculty,Shenyang Jianzhu University,Shenyang Liaoning 110159,China;School of Artificial Intelligence,Shenyang University of Technology Shenyang Liaoning 110000,China)
出处
《计算机仿真》
北大核心
2022年第4期161-165,共5页
Computer Simulation
基金
国防科技创新特区项目(18-H863-00-TS-002-034-01)。
关键词
平滑化
区域填充
水下图像
倒影去除
倒影区域识别
水下图像增强
Smoothing
Area filling
Underwater image
Reflection removal
Reflection area recognition
Underwater image enhancement
