Explorations on Object-Oriented Classification for Ground Targets from High-Resolution Image

Under the constant advancement of remote sensing technology, more and more high spatial resolution data are becoming available, the developments not only greatly increased our capability and accuracy in information extraction but also opened new application opportunities. Existing research activities have focused on how to increase the utilization efficiency of high resolution data, particularly in a wide range of applications such as urban environment, urban-planning, precision agriculture, forest inventory and risk assessment, however, the overall level of automation is still low. The paper adopted the multi-feature information of high-resolution images, objected-oriented image analysis approaches and multi-scale image segmentation technology, established a corresponding relevant knowledge base to extract information. Also the results show the new approach provided a satisfying solution to improve the level of autoimmunization and intelligentization of high-resolution remote sensing data process and application.