本发明提出了一种基于YOLOv8的绝缘子缺陷类型检测方法，基于DRB和DWR思路本发明提出一种全新的DRR模块，将其与YOLOv8模型的C2f模块结合后，加强了模型对于不同尺度目标的检测效果，有效的提高了缺陷检测的精度；同时，根据设计的网络模型将测试集输入评估模型输出绝缘子缺陷检测结果，从而能够面向多种绝缘子的多类缺陷在...
C2f_RFAConv, C3_RFCBAMConv, C2f_RFCBAMConv, C3_RFCAConv, C2f_RFCAConv, C3_FocusedLinearAttention, C2f_FocusedLinearAttention, C3_AKConv, C2f_AKConv, AKConv, C3_MLCA, C2f_MLCA, C3_UniRepLKNetBlock, C2f_UniRepLKNetBlock, C3_DRB, C2f_DRB, C3_DWR_DRB, C2f_DWR_DRB, CSP_EDLA...
C2f_RFAConv, C3_RFCBAMConv, C2f_RFCBAMConv, C3_RFCAConv, C2f_RFCAConv, C3_FocusedLinearAttention, C2f_FocusedLinearAttention, C3_AKConv, C2f_AKConv, AKConv, C3_MLCA, C2f_MLCA, C3_UniRepLKNetBlock, C2f_UniRepLKNetBlock, C3_DRB, C2f_DRB, C3_DWR_DRB, C2f_DWR_DRB, CSP_EDLA...
基于YOLOv8n的模型,在C2f中引入采用DilatedReparamBlock修改的DWR模块,并在主干网络上加入MLCA注意力机制.使用改进后模型检测的实验数据表明,与原模型YOLOv8n相比,YOLOv8nC2f_DWR_DRB-MLCA模型的mAP@0.5和mAP@0.5-0.95分别为88.1%和39.5%,对比原模型分别提高了4.1%和1.6%,有效提升了对驾驶员在驾驶过程中因疲劳...
本发明提出了一种基于YOLOv8的绝缘子缺陷类型检测方法,基于DRB和DWR思路本发明提出一种全新的DRR模块,将其与YOLOv8模型的C2f模块结合后,加强了模型对于不同尺度目标的检测效果,有效的提高了缺陷检测的精度;同时,根据设计的网络模型将测试集输入评估模型输出绝缘子缺陷检测结果,从而能够面向多种绝缘子的多类缺陷在同一...
C2f_RFAConv, C3_RFCBAMConv, C2f_RFCBAMConv, C3_RFCAConv, C2f_RFCAConv, C3_FocusedLinearAttention, C2f_FocusedLinearAttention, C3_AKConv, C2f_AKConv, AKConv, C3_MLCA, C2f_MLCA, C3_UniRepLKNetBlock, C2f_UniRepLKNetBlock, C3_DRB, C2f_DRB, C3_DWR_DRB, C2f_DWR_DRB, CSP_EDLA...