Related Experiment Video
Updated: May 21, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
442
A novel hybrid deep learning approach for super-resolution and objects detection in remote sensing
Muhammad Asif1, Mohammad Abrar2, Faizan Ullah1
1Department of Computer Science, Bacha Khan University, Charsadda, 24420, Pakistan.
Scientific Reports
|May 17, 2025
Summary
This study introduces a novel object detection framework for remote sensing, enhancing image resolution and feature extraction to improve accuracy in complex scenarios. The hybrid model significantly boosts detection performance, offering a robust solution for critical applications.
Area of Science:
- Remote Sensing
- Computer Vision
- Artificial Intelligence
Background:
- Object detection in remote sensing is challenged by low resolution, complex backgrounds, occlusions, and scale variations.
- These challenges are critical for applications like disaster response, environmental monitoring, and surveillance.
Purpose of the Study:
- To propose a robust object detection framework for remote sensing images.
- To integrate super-resolution techniques with advanced feature extraction for improved detection accuracy.
Main Methods:
- A hybrid model combining Advanced StyleGAN for super-resolution and Swin Transformer for feature extraction was developed.
- Data augmentation and preprocessing techniques were employed to enhance the training dataset diversity and accuracy.
Main Results:
- The framework achieved high performance metrics: mAP@0.5 of 97.2%, mAP@0.5:0.95 of 72.8%, and F1-Score of 0.93.
- Demonstrated robustness in challenging conditions (low light, fog) and outperformed existing methods like YOLOv9 and DCNN-based approaches.
- Achieved superior detection accuracy and robustness on RSOD and NWPU VHR-10 datasets.
Conclusions:
- The proposed framework represents a significant advancement in remote sensing object detection.
- It offers an effective solution for complex scenarios, improving accuracy and robustness.
- Future work will focus on computational efficiency and expansion to multimodal or dynamic object detection.
