The Study On Dynamic Monitoring Of City Planning Based On ECognition

In order to improve the efficiency of city planning dynamic monitoring,The study extracted the land cover change information from the city area,and introduced theobject-oriented classification software—eCognition,take pixel to pixel comparison and post classification comparison two method to extract the dynamical change information.Experiments show that the resoult of pixel to pixel comparison has a lower precision than post classification comparison,and its scope is limited.The precision of image post classification comparison is affected by the precision of a single image classification accuracy.Its effect better than pixel to pixel comparison,and applied to the most of the images,but it's difficult to find the internal change information.It's shows that the method of object-oriented classification can improve the classification precision effectively, it is the foundation to find land cover change information more rapidly and accurately.Innovation points: It is the first study for extract land cover change information in city planning dynamic monitoring,and for the first time adopt the object-oriented classification software eCognition in the land cover change information extraction.It is to be true that the output of land cover change information.