基于改进 Yolov5 的遥感光伏检测算法

吉林大学学报(信息科学版) ›› 2023, Vol. 41 ›› Issue (5): 801-809.

基于改进 Yolov5 的遥感光伏检测算法

佟喜峰, 杜鑫, 王志宝

东北石油大学计算机与信息技术学院, 黑龙江大庆 163318

收稿日期:2022-09-28 出版日期:2023-10-09 发布日期:2023-10-10
作者简介:佟喜峰(1974— ), 男, 黑龙江大庆人, 东北石油大学副教授, 主要从事图像处理、模式识别研究, (Tel)86-13045496386 (E-mail)csxftong@ 163. com。
基金资助:
黑龙江省自然科学基金资助项目(LH2021F004)

Taget Detection of Photovoltatic Remote Sensing Based on Improved Yolov5 Model

TONG Xifeng, DU Xin, WANG Zhibao

School of Computer and Information Technology, Northeast Petroleum University, Daqing 163318, China

Received:2022-09-28 Online:2023-10-09 Published:2023-10-10

摘要/Abstract

摘要： 针对遥感光伏图像分辨率高、环境噪声较大以及背景复杂等问题, 提出了一种改进 Yolov5 目标检测模型, 以实现对光伏电厂的定位。首先, 在主干特征提取网络的卷积层中添加 CA(Coordinate Attention)坐标注意力机制提高网络特征的学习能力; 其次, 将 Ghostconv 网络结构加入到 Backbone 中, 用 Ghostconv 网络模块替换 Conv 网络模块; 设计新的 GhostC3 网络代替原来的 C3 网络模块, 提高模型的学习效率; 最后, 将损失函数由 GIoU_Loss 函数改为 SIoU_Loss 函数。实验结果表明, 相比原 Yolov5 方法, 改进算法的平均精度均值 mAP、精准率和召回率分别达到了 97. 5% 、98. 9% 和 94. 9% , 提升了 1. 8% 、1. 7% 和 5. 8% , 验证了该算法对光伏检测具有很好的效果。

关键词: 光伏, 遥感图像, 目标检测, Yolov5 模型

Abstract: Taget Detection of Photovoltatic Remote Sensing Based on Improved Yolov5 Model TONG Xifeng, DU Xin, WANG Zhibao (School of Computer and Information Technology, Northeast Petroleum University, Daqing 163318, China) Abstract: Aiming at high-sensing photovoltaic image resolution, high environmental noise, and complex background, an improved Yolov5 model is proposed to achieve positioning of photovoltaic power plants. First of all, the CA(Coordinate Attention) mechanism is added to the compassionate layer of the main feature extraction network to improve the learning ability of the network characteristics; second, the Ghostconv network structure is added to Backbone, useing the Ghostconv network module to replace the Conv network module, designing a new GhostC3 network network instead of the original C3 network module to improve the learning efficiency of the model; finally, the GIoU_Loss function is changed to the SIoU_Loss function. Compared with the original Yolov5 method, the average accuracy of the improved algorithm mAP, accuracy, and recall rate reached 97. 5% , 98. 9% , and 94. 9% , respectively, which have increased by 1. 8% , 1. 7% , and 5. 8% , respectively. The algorithm has a good effect on photovoltaic detection.

Key words: photovoltaic, remote sensing images, target detection, Yolov5

中图分类号:

TP753

佟喜峰, 杜鑫, 王志宝. 基于改进 Yolov5 的遥感光伏检测算法[J]. 吉林大学学报(信息科学版), 2023, 41(5): 801-809.

TONG Xifeng, DU Xin, WANG Zhibao . Taget Detection of Photovoltatic Remote Sensing Based on Improved Yolov5 Model[J]. Journal of Jilin University (Information Science Edition), 2023, 41(5): 801-809.