科技通报
科技通報
과기통보
Bulletin of Science and Technology
2015年
10期
208-210
,共3页
角点检测%遥感识别%地理特征
角點檢測%遙感識彆%地理特徵
각점검측%요감식별%지리특정
corner detection%remote sensing recognition%geographic characteristics
在不同光照和旋转角度下进行地理特征遥感识别是测绘领域的难点,为了实现对地理地形的准确测绘,需要进行地理特征遥感识别优化设计.传统方法中,采用遥感图像边缘特征检测算法,当遥感图像出现旋转角度时,识别性能不好.针对这一问题,提出一种角点检测的地理特征遥感识别算法,并进行系统设计与实现,系统通过图像传感模块采集遥感图像,经过角点特征处理器提取图像特征.在不同光照和旋转角度,采用单尺度Harris角点检测算法对图像进行检测并提取角点,得到地理遥感特征数据采集的角点模型,基于角点检测,进行遥感识别算法改进和系统实现.实验结果表明,采用该算法和系统,地理特征遥感识别准确率高,耗时较少,提高对地理特征遥感识别性能.为实现对地理地形的准确测绘提供依据.
在不同光照和鏇轉角度下進行地理特徵遙感識彆是測繪領域的難點,為瞭實現對地理地形的準確測繪,需要進行地理特徵遙感識彆優化設計.傳統方法中,採用遙感圖像邊緣特徵檢測算法,噹遙感圖像齣現鏇轉角度時,識彆性能不好.針對這一問題,提齣一種角點檢測的地理特徵遙感識彆算法,併進行繫統設計與實現,繫統通過圖像傳感模塊採集遙感圖像,經過角點特徵處理器提取圖像特徵.在不同光照和鏇轉角度,採用單呎度Harris角點檢測算法對圖像進行檢測併提取角點,得到地理遙感特徵數據採集的角點模型,基于角點檢測,進行遙感識彆算法改進和繫統實現.實驗結果錶明,採用該算法和繫統,地理特徵遙感識彆準確率高,耗時較少,提高對地理特徵遙感識彆性能.為實現對地理地形的準確測繪提供依據.
재불동광조화선전각도하진행지리특정요감식별시측회영역적난점,위료실현대지리지형적준학측회,수요진행지리특정요감식별우화설계.전통방법중,채용요감도상변연특정검측산법,당요감도상출현선전각도시,식별성능불호.침대저일문제,제출일충각점검측적지리특정요감식별산법,병진행계통설계여실현,계통통과도상전감모괴채집요감도상,경과각점특정처리기제취도상특정.재불동광조화선전각도,채용단척도Harris각점검측산법대도상진행검측병제취각점,득도지리요감특정수거채집적각점모형,기우각점검측,진행요감식별산법개진화계통실현.실험결과표명,채용해산법화계통,지리특정요감식별준학솔고,모시교소,제고대지리특정요감식별성능.위실현대지리지형적준학측회제공의거.
In a different light and rotation angle of remote sensing identification of geographic feature is difficult field of Sur-veying and mapping, in order to achieve accurate mapping of the geographical terrain, the need to optimize the design of geographical features of remote sensing recognition. Put forward the geographical characteristics of a corner detection algo-rithm in remote sensing recognition, and system design and implementation, the system through the image sensing module for acquisition of external image after color processor, image feature extraction. Under different illumination and rotation an-gle, using the single scale Harris corner detection algorithm for detection and extraction of image corner, corner get geogra-phy remote sensing characteristics of data acquisition model based on corner point detection, remote sensing recognition al-gorithm is improved and the system realization. The experimental results show that using this algorithm and system, remote sensing identification geographical features high accuracy, less time consumption, improve on the performance of the geo-graphical features of remote sensing recognition. It can provide the basis for realizing accurate mapping of the geographical terrain.