Research on Pedestrian Detection Model and Compression Technology for UAV Images
Open Access
Abstract
The large view angle and complex background of UAV images bring many difficulties to the detection of small pedestrian targets in images, which are easy to be detected incorrectly or missed. In addition, the object detection models based on deep learning are usually complex and the high computational resource consumption limits the application scenarios. For small pedestrian detection in UAV images, this paper proposes an improved YOLOv5 method to improve the detection ability of pedestrians by introducing a new small object feature detection layer in the feature fusion layer, and experiments show that the improved method can improve the average precision by 4.4%, which effectively improves the pedestrian detection effect. To address the problem of high computational resource consumption, the model is compressed using channel pruning technology to reduce the consumption of video memory and computing power in the inference process. Experiments show that the model can be compressed to 11.2 MB and the GFLOPs of the model are reduced by 11.9% compared with that before compression under the condition of constant inference accuracy, which is significant for the deployment and application of the model.Keywords
Funding Information
- National Key Research and Development Program of China (2022YFC3301603)
This publication has 22 references indexed in Scilit:
- Real-Time Small Drones Detection Based on Pruned YOLOv4Sensors, 2021
- The Automatic Detection of Pedestrians under the High-Density Conditions by Deep Learning TechniquesJournal of Advanced Transportation, 2021
- An Improved Lightweight Network MobileNetv3 Based YOLOv3 for Pedestrian Detection2021
- ComNet: Combinational Neural Network for Object Detection in UAV-Borne Thermal ImagesIEEE Transactions on Geoscience and Remote Sensing, 2020
- UAV-YOLO: Small Object Detection on Unmanned Aerial Vehicle PerspectiveSensors, 2020
- Path Aggregation Network for Instance Segmentation2018
- Learning Efficient Convolutional Networks through Network Slimming2017
- You Only Look Once: Unified, Real-Time Object Detection2016
- Microsoft COCO: Common Objects in Context2014
- Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition2014