Automated object recognition in high-resolution optical remote sensing imagery

Xinhao Cai

Abstract:

With the rapid development of remote sensing technology, high-quality remote sensing images have become widely accessible. The automated object detection and recognition of these images, which aims to automatically locate objects of interest in remote sensing images and distinguish their specific categories, is an important fundamental task in the field. It provides an effective means for geo-spatial object monitoring in many social applications, such as intelligent transportation, urban planning, environmental monitoring, and homeland security.
Actually, automated interpretation of remote sensing data is a challenging task due to the wide space coverage and complicated image background. Traditional methods mainly focus on manually designed features to represent objects of interest, which may limit the model's performance. Recently, with the growing wave of deep learning, methods based on the convolutional neural network (CNN) have made great progress in this field. However, it still has difficulty in some specific tasks such as densely distributed small object detection, and requires highly advanced techniques.