Abstract:
Object detection is a crucial remote sensing task. Due to multiscale and multiclass object scales, intraclass differences, and interclass similarity, geospatial object detection remains difficult. This letter proposes an end-to-end feature-reflowing pyramid network (FRPNet) for these issues. Two FRPNet advantages improve object detection accuracy. We embed a nonlocal block into the backbone to determine relevancy between geospatial image regions for discriminative features. A feature-reflowing pyramid structure fuses fine-grained features from the adjacent lower level to generate high-quality feature presentation for each scale, improving multiscale and multiclass object detection. FRPNet outperforms several state-of-the-art detection methods in mean average precision (mAP) on a public remote sensing data set DIOR.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
