【摘要】 SPOT5图像的空间分辨率高,局部异质性较大,采用基于像元的传统方法分类精度低,难以满足实际应用的需要。以北京市海淀区SPOT5图像为例,应用面向对象方法对其进行分类试验,并将该方法与传统基于像元方法的分类结果进行对比分析。结果表明:利用面向对象方法对SPOT5遥感图像进行分类,不仅使分类结果具有丰富的语义信息,有效抑制“椒盐现象”的发生,还可以显著提高分类精度。更多还原

【Abstract】 SPOT5 image is widely used in urban planning,investigation of land utilization,environmental management,public security etc.for its relatively high-resolution and cheap price.Classical classification approaches based on pixels have a low overall accuracy and can not satisfy the application demand in reality due to SPOT5 image having higher resolution and more local heterogeneity.In this paper,object-oriented approach is introduced into SPOT5 image classification.And a general approach and workflow are illustrated on applications of object-oriented approach for high-resolution image classification.Taking Haidian District,Beijing City as the test area,a case study on SPOT5 image classification with object-oriented approach is carried out.In order to verify the accuracy of object-oriented classification,a comparison between this approach and classical classification approaches has been carried out.The case study shows that the application of object-oriented approach on SPOT5 image classification not only can have more semantic information,reduce the"Pepper and Salt Phenomenon"effectively,but also can improve the overall classification accuracy of SPOT5 image.更多还原