High-resolution remote sensing scene target detection method based on size balance FCOS

Abstract

The invention discloses a high-resolution remote sensing scene target detection method based on a size balance FCOS. The method comprises the steps: carrying out the dynamic adjustment of a centrality coefficient according to the regression information of each target through the centrality and frame regression stage of a size balance coefficient in a target detection module of the FCOS, distributing a reasonable weight for the frame regression process of each positive sample, using a high-resolution remote sensing target to detect a remote sensing data set for model training, and using the model for remote sensing ground features for recognition. According to the method, the defects of targets of different sizes in an FCOS centrality evaluation system are fully considered, loss weight reinforcement is carried out on samples, distributed at the edge, of positive samples in a small target anchor frame, redundancy loss contribution in a large target is inhibited, and target size balance is achieved; and the size balance FCOS improves the precision of target detection in a high-resolution remote sensing scene under the condition of not introducing extra overhead to a model reasoning stage.

Description

technical field [0001] The invention belongs to the technical field of computer vision and remote sensing image application, and in particular relates to a high-resolution remote sensing scene target detection method based on size-balanced FCOS. Background technique [0002] In recent years, satellite technology has achieved rapid development, and the application fields of remote sensing images have also been continuously expanded, especially in the fields of meteorology, geology, agriculture, forestry, military affairs, and smart cities. Through remote sensing detection, it is possible to conduct multi-level, perspective, and time observations of large-scale regional images on the earth in a short period of time, which is an important means to obtain environmental information and earth resources; through deep learning-based target detection technology, it can be efficient and accurate Realize the extraction of high-resolution remote sensing image features and reduce the cos...

AI Technical Summary

Problems solved by technology

[0006] The existing FCOS assigns different weights to different positive sample points through the centrality method, but the centrality cannot correctly reflect the quality of the anchor frame regression, especially in the case of remote sensing scenes dominated by small targets

For small objects, the centrality weight assignment strategy has a higher probability, so that the positive sample points of small objects fall on the edge of the anchor box of small objects, resulting in the positive sample points mistakenly obtaining low weights; for large objects, there will be redundancy The positive samples contribute low-quality regression loss, which affects the final target detection

Images