Research Article
Improving Vision Clarity and Object Detection Accuracy in Heavy Rain Base on Neural Network
@INPROCEEDINGS{10.1007/978-3-031-20398-5_9, author={Chi Han Chen and Chien Hung Wu and Cheng Jun Wu and Rung Shiang Cheng}, title={Improving Vision Clarity and Object Detection Accuracy in Heavy Rain Base on Neural Network}, proceedings={Smart Grid and Internet of Things. 5th EAI International Conference, SGIoT 2021, Virtual Event, December 18-19, 2021, Proceedings}, proceedings_a={SGIOT}, year={2022}, month={11}, keywords={Generative Adversarial Network(GAN) Intersection over U ion(IoU) Non-Maximum suppression Object detection Self-driving car}, doi={10.1007/978-3-031-20398-5_9} }- Chi Han Chen
Chien Hung Wu
Cheng Jun Wu
Rung Shiang Cheng
Year: 2022
Improving Vision Clarity and Object Detection Accuracy in Heavy Rain Base on Neural Network
SGIOT
Springer
DOI: 10.1007/978-3-031-20398-5_9
Abstract
In heavy rain situation, clarity of both human vision and computer vision are significantly reduced. Rain removal GAN network is developed to resolve this problem. However, we find that this kind of network causes the decreasing of detection accuracy. In this work, we analysis the performance and the propose a detection network to get higher accuracy. Through the comparison of experimental results, our method can improve the IOU accuracy and detect confidence.
Copyright © 2021–2025 ICST
