无锡九方科技有限公司,江苏无锡 214011
[ "许立兵 (1988-),男,江苏盐城人,硕士,高级工程师,主要从事大数据处理、机器学习的研究。E-mail: jndxxlb@163.com" ]
纸质出版日期:2023-12-15,
移动端阅览
许立兵. 基于YOLOv7的动态图像定位与识别[J]. 新一代信息技术, 2023, 6(23): 01-08
XU Li-bing. Dynamic Image Location and Recognition Based on YOLOv7[J]. New Generation of Information Technology, 2023, 6(23): 01-08
许立兵. 基于YOLOv7的动态图像定位与识别[J]. 新一代信息技术, 2023, 6(23): 01-08 DOI: 10.3969/j.issn.2096-6091.2023.23.001.
XU Li-bing. Dynamic Image Location and Recognition Based on YOLOv7[J]. New Generation of Information Technology, 2023, 6(23): 01-08 DOI: 10.3969/j.issn.2096-6091.2023.23.001.
目前行人检测技术已趋于成熟,但动态图像检测依然是难点,尤其是在动态图像中需要快速对图像定位与识别的技术。针对动态图像定位与识别困难的问题,本文提出了一种基于YOLOv7的快速定位与识别方法。该方法不同于其他实时目标检测方法集中在高效体系结构设计的思路,YOLOv7将优化重点放在训练过程优化上,通过引入bag-of-freebies模块训练方法,采用增加训练成本但不增加推理成本的方式来提高目标定位与识别的准确性。实验表明,该方法能有效减少实时目标检测器40%左右的参数和50%的计算量,对于动态图像有更好的定位与识别效果。
Despite pedestrian detection technology has become mature
dynamic image detection is still difficult
especially in the dynamic image
which needs fast image positioning and recognition. Aiming at the difficulty of dynamic image location and recognition
this paper proposed a fast location and recognition method based on YOLOv7. This method is different from other real-time target detection methods
which mainly focus on the efficient architecture design. YOLOv7 focuses on the optimization of the training process
and improves the accuracy of target location and recognition by introducing the bag-of-freebies module training method
which increases the training cost but does not increase the reasoning cost. The experiment shows that the method can effectively reduce the parameters of the real-time target detector by about 40% and the calculation amount by 50%
and has better positioning and recognition effect for dynamic images.
YOLOv7动态图像目标定位与识别bag-of-freebies
YOLOv7dynamic imagetarget location and identificationbag-of-freebies
WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2023: 7464-7475.
YANG F X, ZHANG X, LIU B Q. Video object tracking based on YOLOv7 and DeepSORT[EB/OL]. (2022-07-25)[2024-04-08]. https://arxiv.org/pdf/2207.12202https://arxiv.org/pdf/2207.12202.
HUSSAIN M, AL-AQRABI H, MUNAWAR M, et al. Domain feature mapping with YOLOv7 for automated edge-based pallet racking inspections[J]. Sensors, 2022, 22(18): 6927.
石争浩, 仵晨伟, 李成建, 等. 航空遥感图像深度学习目标检测技术研究进展[J]. 中国图象图形学报, 2023, 28(9): 2616-2643.
万琴, 李智, 李伊康, 等. 基于改进YOLOX的移动机器人目标跟随方法[J]. 自动化学报, 2023, 49(7): 1558-1572.
陈韦学, 朱猛, 刘志成, 等. 基于YOLOv4的红外多波段图像目标检测算法[J]. 计算机与数字工程, 2023, 51(9): 2038-2042.
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). Piscataway: IEEE, 2016: 779-788.
MASUREKAR O, JADHAV O, KULKARNI P, et al. Real time object detection using YOLOv3[J]. International Research Journal of Engineering and Technology (IRJET), 2020, 7(3): 3764-3768.
WANG Y, WANG H Y, XIN Z H. Efficient detection model of steel strip surface defects based on YOLO-V7[J]. IEEE Access, 2022, 10: 133936-133944.
GE Z, LIU S, WANG F, et al. Yolox: Exceeding yolo series in 2021[EB/OL]. (2021-07-18)[2024-04-08]. https://arxiv.org/pdf/2107.08430https://arxiv.org/pdf/2107.08430.
JJIANG K L, XIE T Y, YAN R, et al. An attention mechanism-improved YOLOv7 object detection algorithm for hemp duck count estimation[J]. Agriculture, 2022, 12(10): 1659.
贾可心, 马正华, 朱蓉, 等. 注意力机制改进轻量SSD模型的海面小目标检测[J]. 中国图象图形学报, 2022, 27(4): 1161-1175.
赵加坤, 孙俊, 韩睿, 等. 基于改进的Faster Rcnn遥感图像目标检测[J]. 计算机应用与软件, 2022, 39(5): 192-196.
谷永立, 宗欣欣. 基于深度学习的目标检测研究综述[J]. 现代信息科技, 2022, 6(11): 76-81.
戚玲珑, 高建瓴. 基于改进YOLOv7的小目标检测[J]. 计算机工程, 2023, 49(1): 41-48.
CHEN C, LIU M Y, TUZEL O, et al. R-CNN for small object detection[C]//Proceedings of IEEE International Conference on Computer Vision. Piscataway: IEEE, 2016: 214-230.
KRISHNA H, JAWAHAR C V. Improving small object detection[C]//2017 4th IAPR Asian Conference on Pattern Recognition (ACPR). Piscataway: IEEE, 2017: 340-345.
ZHANG W, WANG S H, THACHAN S, et al. Deconv R-CNN for small object detection on remote sensing images[C]//IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium. Piscataway: IEEE, 2018: 2483-2486.
HAN J M, DING J, XUE N, et al. ReDet: A rotation-equivariant detector for aerial object detection[C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2021: 2785-2794.
BENJUMEA A, TEETI I, CUZZOLIN F, et al. YOLO-Z: Improving small object detection in YOLOv5 for autonomous vehicles[EB/OL]. (2021-12-22)[2024-04-08]. https://arxiv.org/pdf/2112.11798https://arxiv.org/pdf/2112.11798.
FU H X, SONG G Q, WANG Y C. Improved YOLOv4 marine target detection combined with CBAM[J]. Symmetry, 2021, 13(4): 623.
GEVORGYAN Z. SIoU loss: More powerful learning for bounding box regression[EB/OL]. (2022-05-25)[2024-04-08]. https://arxiv.org/pdf/2112.11798https://arxiv.org/pdf/2112.11798.
0
浏览量
17
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构