基于Faster RCNN的镁还原罐工人检测算法
|
摘要
针对金属镁冶炼还原罐排渣机器人在工作时需要对还原罐工人进行有效规避的问题,提出一种改进的Faster RCNN镁还原罐工人检测方法,使用多层卷积神经网络替代传统工业检测方法。运用深度学习目前主要的目标检测框架Faster RCNN,以其算法为基础,对其中的RPN网络进行改进并提出一种"RPN-Incep"结构,解决还原罐工人的检测问题。同时针对提取特征分辨率小的问题,提出特征层堆叠技术,将多个卷积层同时堆叠输入,增强对还原罐工人的检测性能。实验对比表明,改进的Faster RCNN可以解决还原罐工人的规避问题,在还原罐工人数据集上检测识别率可以到达90%以上,并在公开数据集Caltech上对算法进行了验证。
For tank after magnesium′s smelting slag discharge robots at work effectively for the reduction pot workers need to avoid the problem,this paper puts forward an improved Faster-RCNN magnesium reduction pot workers detection method,using multiple convolution neural network instead of traditional industrial detection method.Faster RCNN,the mainstream target detection framework of deep learning,was used to improve the RPN network based on its algorithm and put forward a " RPN-Incep" structure to solve the detection problem of restitution tank workers.At the same time,in order to solve the problem of low resolution of extraction features,a feature layer stack technique is proposed,which can stack multiple convolution layers simultaneously to enhance the detection performance of workers in small target reduction tanks.The experimental comparison shows that the improved Faster RCNN can solve the avoidance problem of the regenerator workers,and the detection and identification rate can reach more than 90%in the data set of the regenerator workers.The algorithm is validated on Caltech.
For tank after magnesium′s smelting slag discharge robots at work effectively for the reduction pot workers need to avoid the problem,this paper puts forward an improved Faster-RCNN magnesium reduction pot workers detection method,using multiple convolution neural network instead of traditional industrial detection method.Faster RCNN,the mainstream target detection framework of deep learning,was used to improve the RPN network based on its algorithm and put forward a " RPN-Incep" structure to solve the detection problem of restitution tank workers.At the same time,in order to solve the problem of low resolution of extraction features,a feature layer stack technique is proposed,which can stack multiple convolution layers simultaneously to enhance the detection performance of workers in small target reduction tanks.The experimental comparison shows that the improved Faster RCNN can solve the avoidance problem of the regenerator workers,and the detection and identification rate can reach more than 90%in the data set of the regenerator workers.The algorithm is validated on Caltech.
引文
[1]程广涛,陈雪,郭照庄.基于HOG特征的行人视觉检测方法[J].传感器与微系统,2011,30(7):68-70,74.
[2]徐渊,许晓亮,李才年,等.结合SVM分类器与HOG特征提取的行人检测[J].计算机工程,2016,42(1):56-60,65.
[3]芮挺,费建超,周遊,等.基于深度卷积神经网络的行人检测[J].计算机工程与应用,2016,52(13):162-166.
[4]HE K M,ZHANG X Y,REN S Q,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[C].Computer Vision-ECCV,2014.
[5]REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:Unified,real-time object detection[C].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2016:779-788.
[6]LIU W,ANGUELOV D,ERHAN D,et al.SSD:Single shot multibox detector[C].European Conference on Computer Vision,Springer,2016:21-37.
[7]REN S,HE K,GIRSHICK R,et al.Faster R-CNN:Towards real-time object detection with region proposal networks[C].29th Annual Conference on Neural Information Processing Systems,2015:91-99.
[8]EVERINGHAM M,GOOL L,WILLIAMS C K I,et al.The pascal visual object classes(VOC)challenge[J].International Journal of Computer Vision,2010,88(2):303-338.
[9]LONG J,SHELHAMER E,DARRELL T.Fully convolutional networks for semantic segmentation[C].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2015:3431-3440.
[10]GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich feature hierarchies for accurate object detection and semantic segmenta-tion[C].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2014:580-587.
[11]GIRSHICK R.Fast R-CNN[C].Proceedings of the IEEE International Conference on Computer Vision,2015:1440-1448.
[12]UIJLINGS J R R,VAN DE SANDE K E A,GEVERS T,et al.Selective search for object recognition[J].International Journal Of Computer Vision,2013,104(2):154-171.
[13]ZITNICK C L,DOLLR P.Edge boxes:Locating object proposals from edges[C].13th European Conference on Computer Vision,2014:391-405.
[14]SZEGEDY C,LIU W,JIA Y Q,et al.Going deeper with convolutions[J].Computer Vision and Pattern Recognition,2014,arXiv:1409.4842.
[15]ZEILER M D,FERGUS R.Visualizing and understanding convolutional networks[C].European Conference on Computer Vision,Springer,2014:818-833.
[16]SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[J].Computer Vision and Pattern Recognition,2014,arXiv:1409.1556.
[17]疏国会,欧阳一鸣,苏本跃.面向监视视频实时分析的快速行人检测方法[J].计算机应用研究,2017,34(4):1257-1260.
[18]DALAL N,TRIGGS B.Histograms of oriented gradients for human detection[C].IEEE Computer Society Conference on Computer Vision and Pattern Recognition,2005:886-893.
[2]徐渊,许晓亮,李才年,等.结合SVM分类器与HOG特征提取的行人检测[J].计算机工程,2016,42(1):56-60,65.
[3]芮挺,费建超,周遊,等.基于深度卷积神经网络的行人检测[J].计算机工程与应用,2016,52(13):162-166.
[4]HE K M,ZHANG X Y,REN S Q,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[C].Computer Vision-ECCV,2014.
[5]REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:Unified,real-time object detection[C].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2016:779-788.
[6]LIU W,ANGUELOV D,ERHAN D,et al.SSD:Single shot multibox detector[C].European Conference on Computer Vision,Springer,2016:21-37.
[7]REN S,HE K,GIRSHICK R,et al.Faster R-CNN:Towards real-time object detection with region proposal networks[C].29th Annual Conference on Neural Information Processing Systems,2015:91-99.
[8]EVERINGHAM M,GOOL L,WILLIAMS C K I,et al.The pascal visual object classes(VOC)challenge[J].International Journal of Computer Vision,2010,88(2):303-338.
[9]LONG J,SHELHAMER E,DARRELL T.Fully convolutional networks for semantic segmentation[C].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2015:3431-3440.
[10]GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich feature hierarchies for accurate object detection and semantic segmenta-tion[C].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2014:580-587.
[11]GIRSHICK R.Fast R-CNN[C].Proceedings of the IEEE International Conference on Computer Vision,2015:1440-1448.
[12]UIJLINGS J R R,VAN DE SANDE K E A,GEVERS T,et al.Selective search for object recognition[J].International Journal Of Computer Vision,2013,104(2):154-171.
[13]ZITNICK C L,DOLLR P.Edge boxes:Locating object proposals from edges[C].13th European Conference on Computer Vision,2014:391-405.
[14]SZEGEDY C,LIU W,JIA Y Q,et al.Going deeper with convolutions[J].Computer Vision and Pattern Recognition,2014,arXiv:1409.4842.
[15]ZEILER M D,FERGUS R.Visualizing and understanding convolutional networks[C].European Conference on Computer Vision,Springer,2014:818-833.
[16]SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[J].Computer Vision and Pattern Recognition,2014,arXiv:1409.1556.
[17]疏国会,欧阳一鸣,苏本跃.面向监视视频实时分析的快速行人检测方法[J].计算机应用研究,2017,34(4):1257-1260.
[18]DALAL N,TRIGGS B.Histograms of oriented gradients for human detection[C].IEEE Computer Society Conference on Computer Vision and Pattern Recognition,2005:886-893.