| 视频卫星数据的运动车辆提取 |
| |
| 引用本文： | 廖一鸣,万剑华,臧文乾,刘文宋,林英豪.视频卫星数据的运动车辆提取[J].测绘科学,2018(4):144-149,174. |
| |
| 作者姓名： | 廖一鸣 万剑华 臧文乾 刘文宋 林英豪 |
| |
| 作者单位： | 中国石油大学(华东),山东青岛266555;中国科学院遥感与数字地球研究所,北京100101
中国石油大学(华东),山东青岛,266555
中国科学院遥感与数字地球研究所,北京,100101
武汉大学测绘遥感信息工程国家重点实验室,武汉,430074 |
| |
| 基金项目： | 高分重大专项课题项目(Y4D00100GF;Y4D0100038)，中科院战略先导专项课题项目(Y1Y02230XD) |
| |
| 摘 要： | 针对新兴的视频卫星数据在道路监控中应用研究较少的问题,该文提出了一种基于特征的视频卫星数据运动车辆提取优化算法。以SkySat-1卫星获取的Las Vegas地区的视频数据为实验数据,采用传统帧差法与阈值法进行运动车辆提取,验证了该方法对视频卫星数据处理的有效性,同时在此基础上根据空间特征信息界定出道路以内的感兴趣区域,进一步根据车辆的几何特征信息对感兴趣区域的疑似车辆目标进行筛选。实验结果表明:所提算法在提取精度和完整度上明显优于基于像素的帧差法与阈值法的提取结果。
|
| 关 键 词： | 视频卫星 SkySat-1 帧差法 阈值法 特征提取 video satellite SkySat-1 frame differential threshold method feature extraction |
Moving vehicle extraction from video satellite data based on characteristics |
| |
| LIAO Yiming,WAN Jianhua,ZANG Wenqian,LIU Wensong,LIN Yinghao.Moving vehicle extraction from video satellite data based on characteristics[J].Science of Surveying and Mapping,2018(4):144-149,174. |
| |
| Authors: | LIAO Yiming WAN Jianhua ZANG Wenqian LIU Wensong LIN Yinghao |
| |
| Abstract: | Aiming at the problem of emerging video satellite data in road monitoring application with few re searches,an optimization algorithm about detection on video satellite data moving vehicles was put forward on the basis of the characteristics.Based on satellite SkySat-1 for the LasVegas area of video data for experimental data,the traditional frame differential method and threshold method were adopted to extract moving vehicles,the effectiveness of the method in video satellite data processing.According to the characters of spatial information,roads within interested regions was defined on the basis was verified.In accordance with the suspected filtered vehicle targets of further interest area,geometrical characteristics of the vehicle information was used.Experimental results showed that the proposed algorithm on the extracting accuracy and integrity was better than frame differential method and threshold method based on pixel extraction. |
| |
| Keywords: | |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
