小班界线的确定是森林资源规划设计调查最基本的内容,它关系到森林分布空间定位的准确性和面积数量的精确性。采用传统的野外对坡勾绘的方法确定小班界线,带有很大的主观性,且外业工作量相当大。在山区,小班边界一般与山谷线、山脊线相吻合,笔者针对山区小班分布的特点,应用地形表面流水数字模拟法,从DEM中提取出山谷线、山脊线,结合SPOT5遥感影像信息,采用图像分割技术自动提取小班边界,并对分割对象进行评价,提取精度最高可达到0.579。
Determining boundaries of sub-compartments is the foundation of forest resource inventory,which related to the accuracy of spatial location of forest distribution and precision of the number of area.The traditional way of the boundaries determination of sub-compartments is based on the method of opposed sloping and drawing on the topographic map,this approach is subjective in some extent,and needs a lot of field surveying works.In a mountain area,boundaries of sub-compartments,ridge and valley axes are coincident in generally,according to this characteristic of forest sub-compartments,this study firstly extracted the valley and ridge axes based on the DEM data by applying hydrologic simulation methodology.Furthermore,with the integration of SPOT5 image,this study applied the image segmentation technique to determine the boundaries of sub-compartments automatically,and assesses the accuracy of extractions,which could be up to 0.579.