Object detection process of domestic high-resolution building remote sensing image based on weak light enhancement
Jinguang Yao, Kun Liu · 2021
Remote sensing images play an important role in acquiring geographic data, acquiring earth resources and emergency disasters. However, the obtained remote sensing images often have some problems, such as low contrast, poor visibility, blur and so on. This paper presents a building object detection process based on weak light enhancement of domestic high-resolution remote sensing images. The whole process mainly includes image enhancement and object detection. Using unsupervised Generative Adversarial Net, image enhancement training can be carried out without low / normal light image pairs, and the enhanced image effect can be detected through object detection algorithm. The experimental results show that the detection process is very helpful to the accuracy and efficiency of building object detection.