基于ModelBuilder的图斑相似度分析

Pattern Similarity Analysis Based on ModelBuilder

在基于高光谱航空影像进行单木树冠分割和分类的过程中,需要选择合理的技术方案对影像分割的结果进行精度评价。基于使用ArcGIS中的空间数据处理工具和ModelBuilder建模工具进行数据处理与分析运算,结合Jaccard相似系数,计算影像分割得到的图斑栅格数据与真实地面验证数据的匹配度;基于ModelBuilder的方法整合了多个空间数据处理和统计分析工具,得出了Jaccard系数,高效地完成了对影像分割结果的精度评价。

In the process of single tree crown segmentation and classification based on hyperspectral aerial image,it is necessary to select a reasonable technical scheme to evaluate the accuracy of image segmentation results.Based on the spatial data processing tools in ArcGIS and ModelBuilder modeling tools for data processing and analysis,combined with Jaccard similarity coefficient,the matching degree of image segmentation and raster data and real ground verification data was calculated.The method based on ModelBuilder integrates multiple spatial data processing and statistical analysis tools,obtains the Jaccard coefficient,and efficiently completes the precision evaluation of image segmentation results.