Page 105 - 无损检测2024年第七期
P. 105

罗 顺,等:
              基于改进 YOLOv8 的针灸用针缺陷检测算法




















                                                  图 6  改进后网络模型检测结果
              17 FPS,因此改进后算法能够满足需求。                               [11]  GUO  N,JIANG  M  Y,GAO  L  J,et  al.  Simam:  A
                                                                     simple, parameter-free attention module for convolutional
              参考文献:                                                  neural networks[C]//International conference on machine
                                                                     learning. PMLR, 2021: 11863-11874.
                [1]  周李洪,龚金科,李兵. 基于Gabor小波的汽车金属材
                                                                  [12]  MA  S  L,XU  Y.MPDIoU:a  loss  for  efficient  and
                   料表面微小缺陷快速识别[J]. 中南大学学报 (自然科学
                                                                     accurate  bounding  box  regression[EB/OL].2023:
                   版),2021,52(4):1099-1108.
                                                                     arXiv, 2307. 07662.
                [2]  周新颖,胡跃明. 基于小波增强与Canny算法融合的
                                                                  [13]  徐晓光,李海. 多尺度特征在YOLO算法中的应用研
                   锂电池极片缺陷检测方法[J]. 电子测量技术,2023,
                                                                     究[J]. 电子测量与仪器学报,2021,35(6):96-101.
                   46(4):149-154.
                                                                  [14]  HE K M,ZHANG X Y,REN S Q,et al.Deep residual
                [3]  SINDAGI V A,SRIVASTAVA S.Domain adaptation
                                                                     learning for image recognition[C]//2016 IEEE Conference
                   for automatic OLED panel defect detection using adaptive
                                                                     on Computer Vision and Pattern Recognition(CVPR).Las
                   support vector data description[J].International Journal
                   of Computer Vision,2017,122(2):193-211.           Vegas,NV,USA: IEEE,2016, 770-778.
                [4]  丛明, 卢长奇,刘冬,等. 基于Refine-ACTDD的铸件外             [15]  祁宣豪,智敏. 图像处理中注意力机制综述[J/OL].
                   观微小缺陷检测方法[J]. 计算机集成制造系统,2022,                     计算机科学与探索,2023:1-20.[2023-06-30].https://
                   28(9):2815-2824.                                  kns.cnki.net/kcms/detail/11. 5602.TP.20230629.
                                                                     1447.002. html.
                [5]  CAO  C  Q,WANG  B,ZHANG  W  R,et  al.An
                   improved  faster  R-CNN  for  small  object  detection[J].    [16]  ZHENG  Z  H,WANG  P,LIU  W,et  al.  Distance-
                   IEEE Access,2019,7:106838-106846.                 IoU  loss:  Faster  and  better  learning  for  bounding  box
                [6]  BAI T B,YANG J W,XU G Y,et al.An optimized      regression[C]//Proceedings of the AAAI conference on
                                                                     artificial intelligence. Ithaca: arXiv, 2020,34(7):12993-
                   railway  fastener  detection  method  based  on  modified
                   Faster R-CNN[J].Measurement,2021,182:109742.      13000.
                [7]  张开生,李昊晨,关凯凯,等. 基于LT-YOLOv5s的                 [17]  ZHANG  Y  F,  REN  W,  ZHANG  Z,  et  al.  Focal
                   PCB缺陷检测方法[J]. 试验室研究与探索,2023,42(2):                and  efficient  IOU  loss  for  accurate  bounding  box
                   108-114.                                          regression[J]. Neurocomputing, 2022, 506: 146-157.
                [8]  REDMON  J,DIVVALA  S,GIRSHICK  R,et  al.     [18]  LI Y X,HOU Q B,ZHENG Z H,et al.Large selective
                   You  only  look  once:unified,real-time  object   kernel  network  for  remote  sensing  object  detection
                   detection[C]//2016  IEEE  Conference  on  Computer   [EB/OL].2023:arXiv:2303. 09030.
                   Vision  and  Pattern  Recognition(CVPR).Las  Vegas,    [19]  HU  J,SHEN  L,SUN  G.Squeeze-and-excitation
                   NV,USA: IEEE,2016:779-788.                        networks[C]//2018  IEEE/CVF  Conference  on
                [9]  LIU  W,ANGUELOV  D,ERHAN  D,et  al.SSD:         Computer  Vision  and  Pattern  Recognition.Salt  Lake
                   single shot MultiBox detector[C]//European Conference   City,UT,USA: IEEE,2018:7132-7141.
                   on Computer Vision.Cham:Springer,2016:21-37.    [20]  OUYANG D L,HE S,ZHANG G Z,et al.Efficient
                [10]  LIN T Y,GOYAL P,GIRSHICK R,et al.Focal loss    multi-scale  attention  module  with  cross-spatial
                   for  dense  object  detection[C]//2017  IEEE  International   learning[C]//ICASSP  2023-2023  IEEE  International
                   Conference on Computer Vision(ICCV).Venice,Italy:   Conference on Acoustics,Speech and Signal Processing
                   IEEE,2017:2999-3007.                             (ICASSP).Rhodes Island,Greece: IEEE,2023:1-5.

                                                                                                          67
                                                                                         2024 年 第 46 卷 第 7 期
                                                                                                  无损检测
   100   101   102   103   104   105   106   107   108   109   110