Optics and Precision Engineering, Volume. 32, Issue 8, 1241(2024)

Dense control valve parts dataset for industrial object detection

Linyi WANG1... Jing BAI1,3,*, Yanmei LI2 and Wenjing LI1 |Show fewer author(s)
Author Affiliations
  • 1Department of Computer Science and Engineering, North Minzu University, Yinchuan75002, China
  • 2Mount Liupan Laboratory, Yinchuan75001, China
  • 3Key Laboratory of Image and Graphics Intelligent Processing of State Ethnic Affairs Commission, North Minzu University, Yinchuan750021, China
  • show less
    Figures & Tables(15)
    PD4CV2023 dataset collection and preprocessing
    PD4CV2023 dataset annotation
    PD4CV2023 dataset amplification
    PD4CV2023 dataset part types
    Example densely obstructed placement of parts
    Example part size differences
    Comparison similar parts
    • Table 1. PD4CV2023 dataset data statistics

      View table
      View in Article

      Table 1. PD4CV2023 dataset data statistics

      图像(张)样本阀体上阀盖套筒柱状阀杆杆状阀杆阀座阀芯上套筒导向套
      训练集40812 0891 8202 3812 2951 9215902 210449195228
      验证集511 52327628330527847260352514
      测试集511 40323228528024059221471722
      总数51015 0152 3282 9492 8802 4396962 691531237264
    • Table 2. Average number of samples per dataset

      View table
      View in Article

      Table 2. Average number of samples per dataset

      数据集平均样本数
      MS COCO7.19
      PASCAL VOC(07++12)2.89
      Open Images5.00
      ImageNet1.37
      DOTA3.50
      LEVIR0.50
      PD4CV202329.00
    • Table 3. Comparison of relevant datasets

      View table
      View in Article

      Table 3. Comparison of relevant datasets

      名称类别图像数用途样本标记数
      MS COCO80123 287一般目标检测886 266
      PASCAL VOC(07++12)2021 503一般目标检测62 199
      ImageNet200349 319一般目标检测478 806
      DOTA152 806航空目标检测188 282
      LEVIR322 000遥感目标检测11 000
      KolektorSDD1399工业缺陷检测52
      DeepPCB63 000电路板缺陷检测8 853
      NEU-CLS61 800钢板缺陷检测1 800
      PD4CV20239510控制阀目标检测15 015
    • Table 4. Comparison of part detection results of different training sets of models on the PD4CV2023 dataset

      View table
      View in Article

      Table 4. Comparison of part detection results of different training sets of models on the PD4CV2023 dataset

      YOLOv5sYOLOv8s
      训练集MS COCOKolektorSDDNEU-CLSPD4CV2023MS COCOKolektorSDDNEU-CLSPD4CV2023
      测试集PD4CV2023PD4CV2023PD4CV2023PD4CV2023PD4CV2023PD4CV2023PD4CV2023PD4CV2023
      mAP00088.2000085.40
    • Table 5. Comparison of detection performance of different object detection methods in the PD4CV2023 dataset

      View table
      View in Article

      Table 5. Comparison of detection performance of different object detection methods in the PD4CV2023 dataset

      年份模型mAPFPS
      2016Faster-RCNN75.3810
      2016SSD75.8323
      2017DETR55.1120
      2018YOLOv380.6717
      2020YOLOv5s88.2047
      2023YOLOv8s85.4042
    • Table 6. Comparison of different object detection methods on parts of different sizes in the PD4CV2023 dataset

      View table
      View in Article

      Table 6. Comparison of different object detection methods on parts of different sizes in the PD4CV2023 dataset

      尺寸类别Faster-RCNNSSDDETRYOLOv3YOLOv5sYOLOv8s均值
      导向套54.1969.8355.1159.4976.8066.4061.86
      较小套筒64.0791.5737.9395.4798.8098.0080.97
      柱状阀杆69.1765.2648.6579.9291.2091.6074.30
      均值62.4875.5547.2378.2988.9385.33
      杆状阀杆68.5183.3257.8582.6991.7087.0078.51
      中等阀座65.2182.4434.0091.4795.5096.3077.49
      阀芯75.2737.9046.4766.3855.2050.7055.32
      均值69.6667.8946.1180.1880.8078.00
      阀体94.3296.5290.8499.0298.6098.5096.30
      较大上阀盖92.3596.5780.4698.4596.3097.1093.54
      上套筒95.3759.0955.3653.1189.5082.9072.56
      均值94.0184.0675.5583.5394.8092.83
    • Table 7. Comparison of different object detection methods on parts of different sizes in the PD4CV2023 dataset

      View table
      View in Article

      Table 7. Comparison of different object detection methods on parts of different sizes in the PD4CV2023 dataset

      规模类别数量(个)比例Faster-RCNNSSDDETRYOLOv3YOLOv5sYOLOv8s均值
      上套筒2371.5895.3759.0955.3653.1189.5082.9072.56
      较小导向套2641.7654.1969.8355.1159.4976.8066.4061.86
      阀芯5313.5475.2737.9046.4766.3855.2050.7055.32
      杆状阀杆6964.6468.5183.3257.8582.6991.7087.0078.51
      均值73.3462.5453.7065.4378.3071.75
      阀体2 32815.5094.3296.5290.8499.0298.6098.5096.30
      柱状阀杆2 43916.2469.1765.2648.6579.9291.2091.6074.30
      较大阀座2 69117.9265.2182.4434.0091.4795.5096.3077.49
      套筒2 88019.1864.0791.5737.9395.4798.8098.0080.97
      上阀盖2 94919.6492.3596.5780.4698.4596.3097.1093.54
      均值77.0286.4758.3892.8796.0896.30
    • Table 8. Comparison of efficiency and error rates of different methods

      View table
      View in Article

      Table 8. Comparison of efficiency and error rates of different methods

      方式分拣速度错误率
      人工50个/h12.00%
      YOLOv5s1 450个/h11.80%
    Tools

    Get Citation

    Copy Citation Text

    Linyi WANG, Jing BAI, Yanmei LI, Wenjing LI. Dense control valve parts dataset for industrial object detection[J]. Optics and Precision Engineering, 2024, 32(8): 1241

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category:

    Received: Sep. 22, 2023

    Accepted: --

    Published Online: May. 29, 2024

    The Author Email: BAI Jing (baijing_nun@163. com)

    DOI:10.37188/OPE.20243208.1241

    Topics