将遥感和地理信息系统技术引入城市规划的实践——运用“神经网络算法解读高精度卫星影像图”指导校园规划

Practice in Introducing Remote Sensing and Geography Information System to Urban Planning——Exerting "Back-Propagated Arithmetic of Neural Networks Unscrambles Shadowgraphs of IKONOS" to Instruct Campus Planning

近年来,运用遥感卫星提供的数据来监测地区环境变化的研究越来越多,并且在许多的规划部门也已经设立了“地理信息数据站”来满足当今规划的需要。为此有必要在遥感数据与规划方案之间架起一座科学合理、稳定可靠的桥梁,即如何解读卫星影像图来科学指导规划部门合理调整区域未来规划发展方向。该文试图运用神经网络算法(BP神经网络算法)解读高精度(IKONOS)卫星影像数据,以此实现对某高校未来规划的指导。作者在文中建立了一座城市环境评价模型,与他人不同的是:该文作者对环境性能评价指标不再是孤立的选自于影像中的数据,而是加入了实地调研所获得的资料作为评价该地区环境优劣的重要参数之一,从而实现将原本评价城市性能好坏的抽象概念进行量化的过程。

In recent years, more and mere remote data has been used In researches to inspecting environment change on earth, Yet many town planning departments also set up ＂data station of geography information system＂ to satisfy today＇s planning needs. Therefore we have the responsibility to bridge logically and reasonably understanding between remote sensing data and town planning, just as how to unscramble secondary planet＇s images to instruct scientifically planning departments to adjust rationally Intending town planning In this paper, the author exerts BP arithmetic to unscramble shadowgraphs of IKONOS, so that to instruct campus planning. The author upbullds an estimate model in town environment, there is difference in my paper that image data has been never exclusive, Investigation on the spot do be one of parameters in estimating the environment＇s ill or not. By this way, we can realize to turn abstract conception on estimating town capability to quantitative one.