由于近岸视频监测技术具有构建成本低、时空分辨率高的特点,近年来已成为海岸动态监测的互补手段.在近岸视频监测中,水边线可作为岸滩边缘位置变化的替代指标,受复杂海滩地形及不规则的波浪及潮汐变化影响,如何从视频图像中准确检测水边线是近岸视频监测所面临的挑战问题之一.本文针对传统图像处理方法在水边线提取中存在的效率不高和抗噪声能力差等问题,将CIELab颜色模型和蚁群优化算法相结合,对台风风暴潮期间石老人海滩的水边线进行提取和定量分析,并与传统算法进行对比.对青岛石老人海滩2011年台风期间的实时影像资料分析结果表明,与传统的提取算法相比,本文提出的方法在数字视频影像的水边线监测应用中可靠性高,并具有良好的细节呈现能力和抗边缘噪声能力,适用于弱边缘水边线的提取.分析结果验证了本方法在极端天气条件下对视频影像中水边线动态变化的自动提取可行性,对构建长时序海滩岸线动态变化影像自动分析系统具有较好的应用价值.%Because of the characteristics of low cost and high spatio-temporal resolution,nearshore video remote sensing technology has become an alternative means for coastal dynamic monitoring in recent years.For nearshore video monitoring,the waterline position can be used as a proxy indicator for mapping the shoreline changes of beach.Under the influence of complex beach terrain and irregular variation of waves and tides,accurate detection of waterline changes from video images has become one of the challenge problems in nearshore video remote sensing.A combined CIELab color model with ant colony optimization algorithm to detect the edge of waterline has been proposed and it has been evaluated under high water level changeduring typhoon storm surge in Shilaoren Beach,Qingdao city.The results of both comparison with traditional methods for edge detection and field images evaluation have showed that the proposed method has better reliability,accuracy and the ability to preserve the detail edges and anti-noise capability,which is particularly suitable for quantifying waterline efficiently.The feasibility of the proposed method for extracting waterline automatically from field video images in extreme weather conditions is demonstrated and it is showed this method is capable to incorporate into an automotive coastal video system for long term shoreline dynamic change monitoring.
展开▼
