[1]郝紫霄,王 琦,高 尚.基于YOLO-v5算法的航拍图像小目标检测改进算法[J].常州大学学报(自然科学版),2023,35(06):45-51.[doi:10.3969/j.issn.2095-0411.2023.06.006]
　HAO Zixiao,WANG Qi,GAO Shang.Improved algorithm for small target detection in aerial images based on YOLO-v5[J].Journal of Changzhou University(Natural Science Edition),2023,35(06):45-51.[doi:10.3969/j.issn.2095-0411.2023.06.006]
点击复制

基于YOLO-v5算法的航拍图像小目标检测改进算法()
分享到：

常州大学学报(自然科学版)[ISSN:2095-0411/CN:32-1822/N]

卷:
第35卷
期数:
2023年06期
页码:
45-51
栏目:
计算机与信息工程：深度学习与目标检测专题
出版日期:
2023-11-28

文章信息/Info

Title:
Improved algorithm for small target detection in aerial images based on YOLO-v5
文章编号:
2095-0411(2023)06-0045-07
作者:
郝紫霄;  王 琦;  高 尚
(江苏科技大学 计算机学院, 江苏 镇江 212100)
Author(s):
HAO Zixiao;  WANG Qi;  GAO Shang
(School of Computer, Jiangsu University of Science and Technology,Zhenjiang 212100, China)
关键词:
航拍图像;  小目标检测;  轻量骨干网络;  注意力机制
Keywords:
aerial image;  small target detection;  lightweight backbone network;  attention mechanism
分类号:
TP 75
DOI:
10.3969/j.issn.2095-0411.2023.06.006
文献标志码:
A
摘要:
航拍图像具有数据量大、目标尺度小而分布稠密的特征,且其视角是俯视,不同于普通图像的平视,因此针对普通图像的传统目标检测算法无法适应航拍图像的目标检测任务。针对航拍图像小目标检测,提出了一种基于YOLO-v5的改进算法Small-Tiny-YOLO-v5。首先,按照 GhostNet网络结构搭建改进算法的骨干网络; 其次,在骨干网络内部加入注意力机制模块; 再次,构建了一个针对微小目标的航拍图像数据集。此外,在改进算法训练阶段融合了迁移学习的思想。实验结果表明,所提改进算法的模型参数量远低于原始YOLO算法; 精度与速度也优于原始算法,在公开数据集与本文构建的数据集中,精度分别提升了0.009和0.024,速度分别提升了73.735%和58.641%。
Abstract:
Aerial image has the characteristics of large amount of data, small target scale and dense distribution, and its angle of view is downcast, which is different from the head up view of ordinary image. Therefore, the traditional target detection algorithm for ordinary image cannot adapt to the target detection task of aerial image. For small target detection in aerial images, an improved algorithm based on YOLO-v5, Small-Tiny-YOLO-v5, was proposed. Firstly, GhostNet network was used as the backbone network of the improved algorithm; secondly, the attention mechanism module was added in the backbone network; thirdly, an aerial image data set for small targets was constructed. In addition, the idea of transfer learning was integrated in the training of the improved algorithm. Experimental results show that the model parameters of the proposed improved algorithm are lower than the original YOLO algorithm, and the accuracy and speed are also better than the original algorithm. In the public dataset and the dataset constructed in this paper, the accuracy has increased by 0.009 and 0.024 respectively, and the speed has increased by 73.735% and 58.641% respectively.

参考文献/References:

[1] 江波, 屈若锟, 李彦冬, 等. 基于深度学习的无人机航拍目标检测研究综述[J]. 航空学报, 2021, 42(4): 524519.
[2] 曾雪. 基于旋转不变梯度方向直方图的航拍图像目标检测[D]. 南京: 东南大学, 2017.
[3] 方路平, 何杭江, 周国民. 目标检测算法研究综述[J]. 计算机工程与应用, 2018, 54(13): 11-18, 33.
[4] GIRSHICK R. Fast R-CNN[C]//2015 IEEE International Conference on Computer Vision(ICCV). Santiago: IEEE, 2015.
[5] LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]//European Conference on Computer Vision. Cham: Springer, 2016: 21-37.
[6] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR). Las Vegas: IEEE, 2016.
[7] 陈丁, 吉哲. 基于改进Faster R-CNN的无人机航拍图像目标检测[J]. 海洋测绘, 2019, 39(5): 51-55.
[8] 魏湧明, 全吉成, 侯宇青阳. 基于YOLO v2的无人机航拍图像定位研究[J]. 激光与光电子学进展, 2017, 54(11): 101-110.
[9] 魏玮, 蒲玮, 刘依. 改进YOLOv3在航拍目标检测中的应用[J]. 计算机工程与应用, 2020, 56(7): 17-23.
[10] 李华清. 基于SSD的航拍图像小目标快速检测算法研究[D]. 西安: 西安电子科技大学, 2018.
[11] LI Z Z. Road aerial object detection based on improved YOLOv5[J]. Journal of Physics: Conference Series, 2022, 2171(1): 012039.
[12] SANDLER M, HOWARD A, ZHU M L, et al. MobileNetV2: inverted residuals and linear bottlenecks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018.
[13] HAN K, WANG Y H, TIAN Q, et al. GhostNet: more features from cheap operations[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). Seattle: IEEE, 2020.
[14] YU F X, LIU C C, WANG D, et al. AntiDote: attention-based dynamic optimization for neural network runtime efficiency[C]//2020 Design, Automation & Test in Europe Conference & Exhibition(DATE). Grenoble: IEEE, 2020.
[15] HU J, SHEN L, ALBANIE S, et al. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8): 2011-2023.
[16] WANG Q L, WU B G, ZHU P F, et al. ECA-net: efficient channel attention for deep convolutional neural networks[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). Seattle: IEEE, 2020.
[17] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//European Conference on Computer Vision. Cham: Springer, 2018: 3-19.
[18] 王洪元, 徐志晨, 陈海琴, 等. 基于金字塔分割和时空注意力的视频行人重识别[J]. 常州大学学报(自然科学版),2023, 35(2): 66-76.
[19] DU D W, QI Y K, YU H Y, et al. The unmanned aerial vehicle benchmark: object detection and tracking[C]//European Conference on Computer Vision. Cham: Springer, 2018: 375-391.
[20] 袁功霖, 尹奎英, 李绮雪. 基于迁移学习的航拍图像车辆目标检测方法研究[J]. 电子测量技术, 2018, 41(22): 77-81.
[21] XIAO Z F, LIU Q, TANG G F, et al. Elliptic Fourier transformation-based histograms of oriented gradients for rotationally invariant object detection in remote-sensing images[J]. International Journal of Remote Sensing, 2015, 36(2): 618-644.

备注/Memo

备注/Memo:
收稿日期: 2023-04-29。
基金项目: 江苏省高等学校基础科学(自然科学)研究面上资助项目(21KJB510028)。
作者简介: 郝紫霄(1998—), 女, 山东滨州人, 硕士生。通信联系人: 王琦(1981—), E-mail: wangqi@just.edu.cn
更新日期/Last Update: 1900-01-01