Multi-resolution Object Detection and Data Fusion for Large-scale Remote Sensing Images based on Deep Learning Method

With the development of aircraft and telemetry satellites, the acquisition of high-resolution remote sensing images becomes easier. Object detection in large-scale remote sensing images has become a valuable problem. At present, the object detector trained in the natural scene has achieved quite good performance, but in the remote sensing scene, the detection result is not unacceptable because the ground object size is too small, dense, and the background is complicated. In order to make the object detector based on deep learning method operating more accurately in large-scale remote sensing images, this paper proposes an algorithm for multi-resolution detection and fusion of data on remote sensing images. The algorithm divides the original large-scale remote sensing image into a plurality of sub-images according to certain parameter settings, and fuses the results after detecting the objects on each sub-image. The process will be adaptively performed multiple times and adopt different resolutions. Finally, all the obtained objects bounding boxes are subjected to non-maximum suppression processing to obtain the final detection result.