计算机与现代化 ›› 2025, Vol. 0 ›› Issue (06): 65-70.doi: 10.3969/j.issn.1006-2475.2025.06.011

• 算法设计与分析 • 上一篇    下一篇

基于深度学习的无人机航拍违建识别算法

  

  1. (西安工程大学计算机科学学院,陕西 西安 710600)
  • 出版日期:2025-06-30 发布日期:2025-07-01
  • 作者简介: 作者简介:孙毅博(2000—),男,河南汝州人,硕士研究生,研究方向:计算机视觉,E-mail: 326530029@qq.com; 牟莉(1972—),女,山东高密人,副教授,硕士,研究方向:智能化信息系统,机器学习,E-mail: 626470481@qq.com。
  • 基金资助:
    基金项目:陕西省科技计划项目(2019CGXNG-015)

Unauthorized Construction Recognition Algorithm of UAV Aerial Photography Based on Deep Learning

  1. (School of Computer Science, Xi'an Polytechnic University, Xi'an 710600, China)
  • Online:2025-06-30 Published:2025-07-01

摘要: 摘要:针对目前传统人工检测违章建筑方法存在漏检、检测效率慢以及高层房屋检测难度高等问题,本文提出一种基于YOLOv5s模型的违建识别算法。首先,在原来框架的Backbone部分加入注意力机制(Coordinate Attention, CA)以提高检测精度;其次,引入双向特征金字塔网络结构(Bidirectional feature pyramid network, BiFPN)加强特征的提取能力,增强模型各个特征层的融合能力;为解决预测框和实际框的角度匹配问题,将损失函数替换为SIoU。实验结果表明:优化后的模型在自制违建数据集上,相比于其他常用模型的性能指标上均有不同程度的提高,精确率P和平均识别准确率mAP分别达到了91.36%、83.45%。改进后的算法提升了性能的同时也保持了较高的运算速度,满足了无人机航拍检测违章建筑的准确性和时效性需求。

关键词: 关键词:违章建筑识别, YOLOv5, 注意力机制, BiFPN, SIoU

Abstract:
Abstract: Considering the issues of missed detection, slow detection efficiency, and high detection difficulty in the current traditional manual method for detecting illegal buildings in high-rise structures, this paper proposes a YOLOv5s-based algorithm for illegal building detection. Firstly, the coordinate attention mechanism is incorporated into the original framework's backbone section to enhance detection accuracy. Secondly, the bidirectional feature pyramid network (BiFPN) structure is introduced to improve both feature extraction ability and fusion capability across different layers of the model. The loss function is replaced with SIoU to address the problem of matching angles between predicted and actual boxes. Experimental results demonstrate that compared to other commonly used models on our self-built dataset, precision P and mAP value reach 91.36% and 83.45%, respectively. The improved algorithm enhances performance while maintaining high computational speed, meeting both accuracy and timeliness requirements for UAV aerial detection of illegal buildings.

Key words: Key words: identification of illegal buildings, YOLOv5, coordinate attention, BiFPN, SIoU

中图分类号: