Targets Recognition In High-Resolution Remote Sensing Image

Due to the constantly increasing coverage and availability of very high resolution remotely sensed data, , there is the possibility to take buildings and airports as automatically recognition targets. It plays an important role in the field of civil navigation and military targets . In recent years, many researchers have done a great deal of work in this area, but many of them are based on low resolution images and the method is still not perfect and mature.The purposed of this paper is to research on automatic target recognition from high-resolution remote-sensing images.In this paper, we first describe some basic theories and methods of target recognition, and then we proposed novel methods for the airport and bridge in the high-resolution remote-sensing images. More specifically, features of this paper can be concluded as follows:1)this paper describes some basic theories and methods of target recognition, including Meanshift segmentation, SURF and texture feature extraction as well as the basic principles of support vector machines.2) An automatic recognition method to identify airport in high-resolution remote-sensing images is proposed which combine SURF keypoints and region Information .The method uses a dual-scale architecture.First,we detect the regions of interest(ROIs) in the coarse scale. Second, we map the ROIs to the original scale. And finally we use prior knowledge to identify the airport.3) An automatic recognition method to identify bridges over water in high-resolution remote-sensing images is proposed. The method is divided into four steps: Bridge crude extraction, calculating the direction of the bridge, bridge accurate extraction and bridge recognition. We use the wheel algorithm to calculate the direction of bridge. Experimental results show that the method not only reduces the amount of computation, but also improves the calculation accuracy.