Research on lightweight object detection network for high-resolution remote sensing images

High resolution remote sensing images contain rich information about surface objects, and detecting key categories of objects in the images is of great significance in the military field. With the rapid development of artificial intelligence technology, image object detection algorithms based on deep learning have become mainstream. This article proposes a lightweight object detection network based on a dual branch structure to address the issue of lightweighting in object detection algorithms. By constructing semantic and spatial branches, it can effectively extract spatial and semantic information of objects. The experimental results on the high-resolution remote sensing image dataset LEVIR show that the method proposed in this paper can maintain high object detection accuracy while significantly reducing computational complexity, and has the foundation for application on low computing devices such as embedded systems.