The research of high resolution image classification based on object-oriented — To Changping District, Beijing City

With the rapid development of aviation and aerospace industry, high spatial resolution data is increasing continuously, using the traditional pixel-based object recognition software cannot make full use of the information contained in images, promoting the development of object-oriented classification techniques. Changping District, Beijing City, as the research object in this paper, using the original image data SPOT5 as data source and object-oriented classification software eCognition8.0 for the appropriate partition parameters setting, choosing good features and attributes to establish the membership function or using the nearest neighbor classifiers achieved the extraction of different objects. The research data shows that the object-oriented classification method has the characteristics of short time-consuming, high-precision and effective.