ALOS卫星遥感数据影像特征分析及应用精度评价

Image Characteristics Analysis and Application Accuracy Assessment of ALOS Data

通过影像灰度等级分析、单变量统计分析和影像纹理特征分析以及平面点位精度、影像分类精度和图斑面积精度评定,针对ALOS数据图像质量开展影像特征研究与应用精度综合评价,并与SPOT5数据对比。ALOS数据在纹理特征及面积精度方面(尤其是较小的图斑面积精度)均明显优于SPOT5;建设用地和农用地二、三级地类的影像分类精度均可达到优良等级;点位中误差精度可满足1∶1万比例尺制图要求。ALOS数据在国土资源调查和管理中具有良好的应用前景,可与SPOT5形成较好的数据互补。

This article is to analyze image characteristics of ALOS data and comprehensively to assess the application accuracies. The methods involve the analysis of gray value features of original image, uni--variant image statistics models of typical sampies, and image texture features are adopted, and also the accuracy assessments for point position of mapping, image classification, and area-classified are implemented. Some different results between ALOS and SPOT5 images are compared. The research shows ALOS has much better accuracy in image texture and more smaller plot area than SPOTS, as to image classification, ALOS also has quite better accuracy for the land use level II and III of construction and agriculture in study area. ALOS can reach the accuracy request for 1：10 000 scale in mapping. Therefore, the conclusion of the research is that ALOS data will have a wide range of prospects of application in the survey and management of land and resources,and will be able to form a replacement with SPOT5 in data sources.