遥感影像植被信息提取及自动区划判读方法研究

Research on Extraction of Vegetation Information from Remote Sensing Image and Method of Automatic Division Interpretation

森林资源是国家重要的自然和战略资源,在维护国土生态安全中具有重要地位。但在森林督查和日常监管中发现非法使用林地和破坏森林资源等涉林违法犯罪问题屡屡发生,仅靠传统的人工目视解译方法,不仅效率低、耗时长、工作量巨大,且存在着人为漏判、误判等情况,为提高森林督查工作效率,有效减少人工差错等情况,提出了一种对于植被区域快速提取并自动检测变化区域的方法。首先对两期影像进行多尺度分割,提取植被指数,通过植被指数阈值进一步提取出基于对象单元的植被区域,将两期植被区域进行对比,即可得到植被增加、减少和未变化区域。

Forest resources are important natural and strategic resources of our country.It plays an important role in maintaining the ecological security of the land,but illegal forest use and destruction of forest resources etc.have been found frequently in forest supervision and daily supervision.However,the traditional manual visual interpretation method,not only inefficient,a long time,but also the huge workload,and there are cases of artificial omission and misjudgment.Therefore,this paper proposes a method for rapid extraction of vegetation areas and automatic detection of changing areas,which is applied to forest supervision.The experiment first performs multi-scale segmentation of the images in the two phase research areas,extracts the vegetation index,and extracts the vegetation areas based on the target units through vegetation index threshold,then the vegetation areas of the two phases are compared to obtain the area of increased vegetation,reduced vegetation,and unchanged vegetation.Practice has proved that this method can greatly improve work efficiency and effectively reduces misjudgment and omission.