摘要
构建了一个基于视频图像的IDL水位新检测模型,可以准确提供监测点积水深度,为监测城市内涝灾情提供科学依据。研究数据为水位视频图像,在ENVI软件中对数据进行校正和重采样。利用能量函数建立水位提取模型并估算水位高度。模型估算的水位和实际水位的统计分析结果为:相关系数的平方R2=0.83,平均绝对误差MAE=1.35 cm,均方根误差RMSE=1.50 cm。
A new algorithm for water level detection is proposed based on video image.The video image data is used as the basic data in this study.And the pro-processing of video image is based on ENVI.The procedure includes correction and resampling.Then the energy function is applied in building the final water level detection model.The water level detection result is validated by the real water level data and the statistical result is R^2 =0.83,MAE =1.35 cm,RMSE =1.50 cm.
出处
《科学技术与工程》
北大核心
2014年第29期114-116,共3页
Science Technology and Engineering
基金
遥感科学国家重点实验室开放基金(OFSLRSS201309)
国家自然科学基金(20110141)资助
