CAJ | 학술논문

由于自然灾害的影响，部分地区的水田缺水无法种植水稻，对水田信息的提取不再只是对种植的水稻提取，因此从高分辨率遥感影像中提取水田信息具有非常重要的作用和意义。本文将水田分为有庄稼水田、地膜水田和未种植水田，利用面向对象的方法对水田信息进行提取。论文在利用eCongniton软件对影像进行多尺度分割，不同的地类采用不同的分割尺度，采用模糊分类的方法，结合影像的光谱、形状、纹理等特征构建水田特征知识库对水田信息进行提取。实验结果表明利用面向对象的方法对水田信息提取有良好的效果。
유우자연재해적영향，부분지구적수전결수무법충식수도，대수전신식적제취불재지시대충식적수도제취，인차종고분변솔요감영상중제취수전신식구유비상중요적작용화의의。본문장수전분위유장가수전、지막수전화미충식수전，이용면향대상적방법대수전신식진행제취。논문재이용eCongniton연건대영상진행다척도분할，불동적지류채용불동적분할척도，채용모호분류적방법，결합영상적광보、형상、문리등특정구건수전특정지식고대수전신식진행제취。실험결과표명이용면향대상적방법대수전신식제취유량호적효과。
Due to the impact of natural disasters, the paddy in some areas can not cultivate rice due to water shortages, so the extraction of information is no longer only just the extraction of rice, thus extracting paddy information from high-resolution remote sensing image has a very important role and significance. This article will divide the paddy into crops paddy, mulch planting paddy and no planting paddy, the object-oriented method is used to extract the information of paddy. This paper uses eCongniton software to conduct multiscale segmentation of image, different land types using different segmentation scale, applies fuzzy classification method to build the paddy feature knowledge base by combining with the spectrum, shapes, textures and other features of images d for paddy information extraction. Experimental results show that the use of object-oriented approach has good results in paddy information extraction.