| Forest vegetation can not only provide the timber and forest products which human needs, but also plays a crucial role to maintain and improve the environment.However, with the progress of human society and the development of economic construction, forest resources are suffering from serious damage. In order to grasp the dynamic changes in forest resources, rational planning of forest systems, need to grasp the real-time and dynamic forest resource data.In recent years, with the rapid development of remote sensing technology and its wide range of applications in forestry, using high-resolution remote sensing images to extract vegetation information has become the world’s hot spots. This article is using the ENVI EX module to extract vegetation information in the QuickBird Image of Jiufeng, then study and compare the classification accuracy between supervised classification and direct output vector of object-oriented classification, explore the accuracy with object-oriented extraction of forest vegetation information. The experiments show that:the object-oriented classification can just set a scale segmentation value to get multi-scale segmentation results in the ENVI EX module, the process is simple and convenient;and by comparing supervised classification and direct vector output of object-oriented classification, we found that the accuracy of former classification is up to90%, obvious higher than the latter a lot which accuracy is only82%, and the latter is easy to automatically generate a lot of very similar but nondescript categories. Therefore, the process that using object-oriented extraction of forest vegetation information on high-resolution remote sensing image, The overall accuracy of training samples for the supervision of was significantly higher than the direct vector output method, it can classify and discriminate vegetation effectively, It provides a reliable foundation and effective method to achieve better extraction of vegetation information. |
PDF Full Text Request