CAJ | 학술논문

标志着中国遥感卫星进入亚米级“高分时代”的“高分二号”（GF-2）卫星在太原成功发射，它将为地质灾害调查提供可靠的数据支持。文章以滑坡灾害频发的云南东川区为研究对象，首先对GF-2卫星采集的研究区数据进行正射校正、配准、融合和裁剪等预处理；然后采用面向对象的分类方法，通过影像分割、合并分块，并结合滑坡灾害的光谱、纹理和数字高程模型等多种信息特征，建立滑坡灾害识别规则，实现滑坡信息的遥感分类；最后对分类结果进行误分对象剔除，得到研究区滑坡信息。结果显示，利用 GF-2卫星数据可以较好地提取滑坡灾害信息，基本满足滑坡的灾害识别要求。文章通过面向对象的滑坡灾害特征提取所输出矢量结果可以方便地计算出滑坡的面积，对滑坡研究提供了极大的便利。
표지착중국요감위성진입아미급“고분시대”적“고분이호”（GF-2）위성재태원성공발사，타장위지질재해조사제공가고적수거지지。문장이활파재해빈발적운남동천구위연구대상，수선대GF-2위성채집적연구구수거진행정사교정、배준、융합화재전등예처리；연후채용면향대상적분류방법，통과영상분할、합병분괴，병결합활파재해적광보、문리화수자고정모형등다충신식특정，건립활파재해식별규칙，실현활파신식적요감분류；최후대분류결과진행오분대상척제，득도연구구활파신식。결과현시，이용 GF-2위성수거가이교호지제취활파재해신식，기본만족활파적재해식별요구。문장통과면향대상적활파재해특정제취소수출시량결과가이방편지계산출활파적면적，대활파연구제공료겁대적편리。
GF-2 satellite was successfully launched from Taiyuan, marking China remote sensing satellite into the submeter age and providing better remote sensing data for the investigation of geological disasters. GF-2 data preprocessing includes orthorectification, registration, fusion and clipping. Combining the spectral characterization, texture and DEM of landslide, this paper uses object-oriented classification method by seg-menting image, merging blocks, building classification rule and eliminating misclassification object to extract landslide information in Dongchuan District, Yunnan Province. The extraction result shows that, the method can meet the landslide recognition requirements with GF-2 data. The vector data extracted from landslide in-formation could be used to calculate landslide area and provide great convenience for landslide research.