Research on Adaptive Classification Algorithm of Remote Sensing Image

With the rapid development of remote sensing technology, the information contained in an image becomes more and more comprehensive. Therefore, how to automatically extract information of interest from image data has become a hot topic. In this paper, the automatic airborne remote sensing image classification algorithm is studied by the use of image pyramid. First, the airborne remote sensing image is divided into several tiles to reduce the amount of data processed at the same time; second, Gaussian smoothing is carried in the studied area in order to weaken the influence of noise; third, image pyramid is established according to the different image scale of the target area; fourth, Canny boundary recognition processing is done in the original image; in the end, a package of stable boundary information is obtained by analyzing the objects' features in different image scale based on the theory "different objects have different performance under different resolutions". Aerial photographs are taken for experiment, the results prove that the algorithm can classify the boundary information effectively. The final classification results can be applied to a series of operations after lined edge procession.