The contact network dropper works in a harsh environment,and suffers from the impact effect of pantographs during running of trains,which may lead to faults such as slack and broken of the dropper wire and broken of t...The contact network dropper works in a harsh environment,and suffers from the impact effect of pantographs during running of trains,which may lead to faults such as slack and broken of the dropper wire and broken of the current-carrying ring.Due to the low intelligence and poor accuracy of the dropper fault detection network,an improved fully convolutional one-stage(FCOS)object detection network was proposed to improve the detection capability of the dropper condition.Firstly,by adjusting the parameterαin the network focus loss function,the problem of positive and negative sample imbalance in the network training process was eliminated.Secondly,the generalized intersection over union(GIoU)calculation was introduced to enhance the network’s ability to recognize the relative spatial positions of the prediction box and the bounding box during the regression calculation.Finally,the improved network was used to detect the status of dropper pictures.The detection speed was 150 sheets per millisecond,and the MAP of different status detection was 0.9512.Through the simulation comparison with other object detection networks,it was proved that the improved FCOS network had advantages in both detection time and accuracy,and could identify the state of dropper accurately.展开更多
为实现桥梁裂缝的快速、准确定位,考虑光照变化、污渍阴影等干扰因素的影响,提出一种结合桥梁检测机和改进单阶段目标检测(you only look once version 3,YOLOv3)算法的桥梁裂缝检测方法。首先,在预处理阶段,采用改进自适应Mask匀光算...为实现桥梁裂缝的快速、准确定位,考虑光照变化、污渍阴影等干扰因素的影响,提出一种结合桥梁检测机和改进单阶段目标检测(you only look once version 3,YOLOv3)算法的桥梁裂缝检测方法。首先,在预处理阶段,采用改进自适应Mask匀光算法对数据集进行处理,矫正阴影和光照不均等问题,提高算法环境适应能力;其次,在目标检测阶段,针对桥梁裂缝的特点,对数据集使用k-means++算法聚类先验框以适应裂缝的不同尺寸,采用广义交并比对YOLOv3损失函数进行改进以提高定位精度;最后,采用迁移学习对YOLOv3进行训练。实验结果表明,在迭代140个epoch后,检测速度可达到31帧/s,平均精度(average precision,AP)达到94.88%,相比于采用原始数据集的原始YOLOv3网络AP值提高了13.16%,能够满足实时性和高精度的检测要求。展开更多
高效和准确的场景文本(efficient and accuracy scene text,EAST)检测算法速度快且结构简单,但是由于文本结构的特殊性,导致在检测中尺寸较小的文本会被遗漏,而较长的文本则完整性较差。针对EAST算法存在的问题提出一种新的自然场景文...高效和准确的场景文本(efficient and accuracy scene text,EAST)检测算法速度快且结构简单,但是由于文本结构的特殊性,导致在检测中尺寸较小的文本会被遗漏,而较长的文本则完整性较差。针对EAST算法存在的问题提出一种新的自然场景文本检测模型。该方法利用自动架构搜索的特征金字塔网络(neural architecture search feature pyramid network,NAS-FPN)设计搜索空间,覆盖所有可能的跨尺度连接提取自然场景图像特征。针对输出层进行修改,一方面通过广义交并比(generalized intersection over union,GIOU)作为指标提升边界框的回归效果;另一方面通过对损失函数进行修改解决类别失衡问题。输出场景图像中任意方向的文本区域检测框。该方法在ICDAR2013和ICDAR2015数据集上都取得了较好的检测结果,与其他文本检测方法相比,检测效果也得到了明显提升。展开更多
针对配网勘灾中人工勘灾效率低下和机巡勘灾需后端分析导致灾情信息反馈不及时的问题,立足于前端实时智能检测模式,提出了基于改进YOLO-ResNet混合神经网络的配网杆塔倾倒实时检测模型。首先,改进传统YOLO-V3的损失函数,利用广义交并比(...针对配网勘灾中人工勘灾效率低下和机巡勘灾需后端分析导致灾情信息反馈不及时的问题,立足于前端实时智能检测模式,提出了基于改进YOLO-ResNet混合神经网络的配网杆塔倾倒实时检测模型。首先,改进传统YOLO-V3的损失函数,利用广义交并比(generalized intersection over union,GIoU)计算目标检测框损失,有效提升杆塔主体检测的准确性。其次,采用ResNet-50定位杆塔端点和中心线,提出一种杆塔姿态判断方法以快速计算杆塔倾斜角度。最后,研发了一种便携式设备并部署了所提模型,以实地采集的数据对模型和设备进行测试,结果表明该设备对杆塔姿态判断的整体准确率达93.48%,设备平均功耗9 W,可用于前端实时智能分析、汇总杆塔受灾情况,验证了模型和设备的有效性。展开更多
基金supported by Natural Science Foundation of Gansu Province(No.20JR10RA216)。
文摘The contact network dropper works in a harsh environment,and suffers from the impact effect of pantographs during running of trains,which may lead to faults such as slack and broken of the dropper wire and broken of the current-carrying ring.Due to the low intelligence and poor accuracy of the dropper fault detection network,an improved fully convolutional one-stage(FCOS)object detection network was proposed to improve the detection capability of the dropper condition.Firstly,by adjusting the parameterαin the network focus loss function,the problem of positive and negative sample imbalance in the network training process was eliminated.Secondly,the generalized intersection over union(GIoU)calculation was introduced to enhance the network’s ability to recognize the relative spatial positions of the prediction box and the bounding box during the regression calculation.Finally,the improved network was used to detect the status of dropper pictures.The detection speed was 150 sheets per millisecond,and the MAP of different status detection was 0.9512.Through the simulation comparison with other object detection networks,it was proved that the improved FCOS network had advantages in both detection time and accuracy,and could identify the state of dropper accurately.
文摘为实现桥梁裂缝的快速、准确定位,考虑光照变化、污渍阴影等干扰因素的影响,提出一种结合桥梁检测机和改进单阶段目标检测(you only look once version 3,YOLOv3)算法的桥梁裂缝检测方法。首先,在预处理阶段,采用改进自适应Mask匀光算法对数据集进行处理,矫正阴影和光照不均等问题,提高算法环境适应能力;其次,在目标检测阶段,针对桥梁裂缝的特点,对数据集使用k-means++算法聚类先验框以适应裂缝的不同尺寸,采用广义交并比对YOLOv3损失函数进行改进以提高定位精度;最后,采用迁移学习对YOLOv3进行训练。实验结果表明,在迭代140个epoch后,检测速度可达到31帧/s,平均精度(average precision,AP)达到94.88%,相比于采用原始数据集的原始YOLOv3网络AP值提高了13.16%,能够满足实时性和高精度的检测要求。
文摘高效和准确的场景文本(efficient and accuracy scene text,EAST)检测算法速度快且结构简单,但是由于文本结构的特殊性,导致在检测中尺寸较小的文本会被遗漏,而较长的文本则完整性较差。针对EAST算法存在的问题提出一种新的自然场景文本检测模型。该方法利用自动架构搜索的特征金字塔网络(neural architecture search feature pyramid network,NAS-FPN)设计搜索空间,覆盖所有可能的跨尺度连接提取自然场景图像特征。针对输出层进行修改,一方面通过广义交并比(generalized intersection over union,GIOU)作为指标提升边界框的回归效果;另一方面通过对损失函数进行修改解决类别失衡问题。输出场景图像中任意方向的文本区域检测框。该方法在ICDAR2013和ICDAR2015数据集上都取得了较好的检测结果,与其他文本检测方法相比,检测效果也得到了明显提升。
文摘针对配网勘灾中人工勘灾效率低下和机巡勘灾需后端分析导致灾情信息反馈不及时的问题,立足于前端实时智能检测模式,提出了基于改进YOLO-ResNet混合神经网络的配网杆塔倾倒实时检测模型。首先,改进传统YOLO-V3的损失函数,利用广义交并比(generalized intersection over union,GIoU)计算目标检测框损失,有效提升杆塔主体检测的准确性。其次,采用ResNet-50定位杆塔端点和中心线,提出一种杆塔姿态判断方法以快速计算杆塔倾斜角度。最后,研发了一种便携式设备并部署了所提模型,以实地采集的数据对模型和设备进行测试,结果表明该设备对杆塔姿态判断的整体准确率达93.48%,设备平均功耗9 W,可用于前端实时智能分析、汇总杆塔受灾情况,验证了模型和设备的有效性。
