论文部分内容阅读
林地信息的获取具有重要意义,可广泛应用于城市规划编制、林地资源分析、水土流失治理等领域。本文基于WorldView-2影像,在研究分割与对象特征的基础上,采用面向对象方法,实现了林地信息提取,其用户精度达到94%,制图精度达到97.2%。然后,将像元二分模型应用于林地类对象,计算其植被覆盖度。试验表明,面向对象方法能高效、准确、经济地提取林地信息;基于林地类对象计算的植被覆盖度能较好地表征植被覆盖情况。
The acquisition of woodland information is of great significance and can be widely used in the fields of city planning, forestland resource analysis, soil erosion control and other fields. Based on the WorldView-2 images, the segmentation of the object and the features of the object are studied, and the object-oriented method is adopted to extract the woodland information. The accuracy of the user is 94% and the accuracy of the drawing is 97.2%. Then, the pixel dichotomy model is applied to forest land class object to calculate its vegetation coverage. Experiments show that object-oriented method can extract woodland information efficiently, accurately and economically. Vegetation coverage calculated based on woodland object can better characterize vegetation cover.