火电厂煤场作业人员的多模态检测方法Multimodal detection method for coal yard operators in thermal power plant
张琨;金坤;张方;唐晓萌;程豪豪;岳益锋;张剑华;
摘要(Abstract):
煤场作业人员的安全防护是火电厂安全有序生产的重要保障.封闭煤场内部场景存在光照条件差、粉尘干扰严重、颜色信息单调、检测范围广的特点,导致在该场景中传统基于可见光图像的人员检测方法无法有效检测人员安全性.本文提出一种基于双光相机的多模态融合深度学习的方法,对封闭煤场作业人员进行精确的目标检测,检测精度较原始检测方法提升了近22%.首先给出了该方法的整体架构,然后阐述具体使用的融合方法及卷积神经网络结构,最后采集真实样本开展了模型训练及测试.实验结果表明,本算法的检测精度高、速度快,实现了煤场恶劣环境条件下作业人员的高效检测.
关键词(KeyWords): 火电厂;煤场;多模态融合;深度学习;卷积神经网络;目标检测
基金项目(Foundation): 中国华电集团有限公司科技项目(CHDKJ17-01-64;CHDKJ16-01-40)
作者(Author): 张琨;金坤;张方;唐晓萌;程豪豪;岳益锋;张剑华;
Email:
DOI:
参考文献(References):
- [1]耿蒲龙,宋建成,刘旭飞,等.基于RGB颜色空间的矿井运动目标检测及跟踪方法[J].太原理工大学学报,2017,48(6):963-968.
- [2]董观利,宋春林.基于视频的矿井行人越界检测系统[J].工矿自动化,2017,43(2):29-34.
- [3]常政威,彭倩,陈缨.基于机器学习和图像识别的电力作业现场安全监督方法[J].中国电力,2020,53(4):155-160.
- [4] Liu W,Anguelov D,Erhan D,et al. SSD:Single shot multibox detector[C]//European Conference on Computer Vision.Amsterdam:Springer, 2016:21-37.
- [5] Redmon J,Divvals S,Girshick R,et al. You only look once:unified realtime object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. IEEE Computer Society. USA:Las Vegas,2016:779-788.
- [6] Ren S,He K M,Girshick R,et al. Faster R-CNN:towards real-time object detection with region proposal networks[C]//Advances in Neural Information Processing Systems.MIT Press,Cambridge,USA,2015:91-99.
- [7] Redmon J,Farhadi A. Yolov3:an incremental improvement[C/OL].[2020-9-16]. https://arxiv.org/pdf/1804.02767.pdf.
- [8] He K M,Zhang X Y,Ren S Q,et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE conference on Computer Vision and Pattern Recognition. Las Vegas:IEEE,2016:770-778.
- [9] Luo L C,Xiong Y H,Liu Y,et al. Adaptive gradient methods with dynamic bound of learning rate[C/OL]//[2020-9-16]. https://arxiv.org/pdf/1902.09843.pdf.