CAJ | 학술논문

农用井等小地物识别是土地整理遥感监测的重要内容之一.针对目前相关系数模板匹配方法效率低的缺点,该文提出一种基于土地整理规划图约束下的灰度归一化相关系数模板匹配方法,对整理区农用井进行识别.选择北京市顺义区赵全营土地整理区2006年QuickBird遥感影像作为基础数据,以TITAN IMAGE二次开发平台作为算法实现环境进行试验.结果表明,该方法在选择合适的农用井模板影像和归一化相关系数阈值后,识别准确率可达88.8%,验证了该文提出的土地整理区农用井遥感识别算法的可行性,同时为土地整理区其他小地物识别提供了一种有效途径.
농용정등소지물식별시토지정리요감감측적중요내용지일.침대목전상관계수모판필배방법효솔저적결점,해문제출일충기우토지정리규화도약속하적회도귀일화상관계수모판필배방법,대정리구농용정진행식별.선택북경시순의구조전영토지정리구2006년QuickBird요감영상작위기출수거,이TITAN IMAGE이차개발평태작위산법실현배경진행시험.결과표명,해방법재선택합괄적농용정모판영상화귀일화상관계수역치후,식별준학솔가체88.8%,험증료해문제출적토지정리구농용정요감식별산법적가행성,동시위토지정리구기타소지물식별제공료일충유효도경.
The recognition of small ground-objects, such as farm wells, is one of the most important contents of the remote sensing monitoring in land consolidation. To resolve the low efficiency shortcoming of the traditional correlation coefficient template matching method, gray normalized correlation coefficient of template matching method was presented, which was constraint-based the land reorganization planning map, to recognize the farm wells in land consolidation. The QuickBird image with high spatial resolution in 2006 was selected as the experimental data, and the study area of land consolidation was located in Zhaoquanying, Shunyi District of Beijing. And the algorithm was implemented in the TITAN IMAGE development environment. The experimental results showed that the recognition accuracy of that method could achieve 88.8% when the template image of the farm wells and the threshold of the normalized correlation coefficient were selected appropriately. The result proves the remote sensing monitoring of land consolidation is feasible. Meanwhile, the study is provides a new effective way to recognition of small objects in the land consolidation projects.