[1]赵文清,孔子旭,赵振兵.隔级融合特征金字塔与CornerNet相结合的小目标检测[J].智能系统学报,2021,16(1):108-116.[doi:10.11992/tis.202004033]
 ZHAO Wenqing,KONG Zixu,ZHAO Zhenbing.Small target detection based on a combination of feature pyramid and CornerNet[J].CAAI Transactions on Intelligent Systems,2021,16(1):108-116.[doi:10.11992/tis.202004033]
点击复制

隔级融合特征金字塔与CornerNet相结合的小目标检测(/HTML)
分享到:

《智能系统学报》[ISSN:1673-4785/CN:23-1538/TP]

卷:
第16卷
期数:
2021年1期
页码:
108-116
栏目:
学术论文—知识工程
出版日期:
2021-01-05

文章信息/Info

Title:
Small target detection based on a combination of feature pyramid and CornerNet
作者:
赵文清13 孔子旭1 赵振兵2
1. 华北电力大学 控制与计算机工程学院,河北 保定 071003;
2. 华北电力大学 电气与电子工程学院,河北 保定 071003;
3. 复杂能源系统智能计算教育部工程研究中心,河北 保定 071003
Author(s):
ZHAO Wenqing13 KONG Zixu1 ZHAO Zhenbing2
1. School of Control and Computer Engineering, North China Electric Power University, Baoding 071003, China;
2. School of Electrical and Electronic Engineering, North China Electric Power University, Baoding 071003, China;
3. Engineering Research Ce
关键词:
CornerNet小目标检测卷积特征图隔级融合上下融合旁路连接特征金字塔
Keywords:
CornerNetsmall target detectionconvolutionfeature mapinterval fusionupper and lower fusionbypass connectionfeature pyramid
分类号:
TP391
DOI:
10.11992/tis.202004033
摘要:
为弥补CornerNet中小目标语义信息弱的缺陷,提出隔级融合特征金字塔的方法,提高小目标平均准确率。 对骨干网络后半部分融合后的4个特征图进行提取,将尺寸较小的特征图进行2次卷积,得到2个新的特征图;运用上下融合、隔级融合和旁路连接的思想,生成融合后的特征图并将其组成特征金字塔。将改进后的算法与当前主流CornerNet、Faster RCNN、RetinaNet算法在MS COCO数据集上进行比较,结果表明,改进后算法在对小目标进行检测时,小目标平均准确率有较大提高。隔级融合特征金字塔在CornerNet上能有效融合高低层特征图,使融合后的特征图有较强的语义信息,提高CornerNet网络的小目标平均准确率。
Abstract:
To improve the problem of the weak semantic information of the small target in CornerNet, a method of the hierarchical fusion feature pyramid is proposed to increase the average accuracy of the small target. The method first extracts the four feature maps after the fusion of the second half of the backbone network, then convolves the feature maps with a smaller size twice to obtain two new feature maps, and finally uses the ideas of the upper and lower fusion, interlevel fusion, and bypass connection to generate a fused feature map and form it into a feature pyramid. The result shows that the average accuracy for small targets obtained by our algorithm has been greatly improved compared with those by current mainstream algorithms, such as CornerNet, Faster RCNN, and RetinaNet on the MS COCO dataset, which demonstrates great superiority. The inter-level fusion feature pyramid can effectively fuse high-level and low-level feature maps on CornerNet, so that the fused feature maps have strong semantic information, and improve the average accuracy of the small targets of the CornerNet network.

参考文献/References:

[1] 赵永强, 饶元, 董世鹏,等. 深度学习目标检测方法综述[J]. 中国图象图形学报, 2020, 25(4):629-654
ZHAO Yongqiang, RAO Yuan, DONG Shipeng, et al. Survey on deep learning object detection[J]. Journal of image and graphics, 2020, 25(4):629-654
[2] 赵文清, 周震东, 翟永杰. 基于反卷积和特征融合的SSD小目标检测算法[J]. 智能系统学报, 2020, 15(2):310-316.
ZHAO Wenqing, ZHOU Zhendong, ZHAI Yongjie. SSD small target detection algorithm based on deconvolution and feature fusion[J]. CAAI transactions on intelligent systems, 2020, 15(2):310-316.
[3] GIRSHICK R. Fast R-CNN[C]//Proceedings of 2015 IEEE International Conference on Computer Vision. Santiago, Chile, 2015:1440-1448.
[4] HE K, ZHANG X, REN S, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE transactions on pattern analysis and machine intelligence, 2015, 37(9):1904-1916.
[5] HE Kaiming, GKIOXARI G, DOLLáR P, et al. Mask R-CNN[C]//Proceedings of 2017 IEEE International Conference on Computer Vision. Venice, Italy, 2017:2961-2969.
[6] 段仲静, 李少波, 胡建军, 等. 深度学习目标检测方法及主流框架综述[J]. 激光与光电子学进展, 2020, 57(12):1-16.
DUAN Zhongjing, LI Shaobo, HU Jianjun,et al. Review of deep learning target detection methods and mainstream frameworks[J]. Progress in laser and optoelectronics, 2020, 57(12):1-16.
[7] LIU Wei, ANGUELOV D, ERHAN D, et al.SSD:single shot multibox detector[C]//Proceedings of 14th European Conference on Computer Vision. Amsterdam, The Netherlands, 2016:21-37.
[8] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once:unified, real-time object detection[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, USA, 2016:779-788.
[9] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedings of 2017 IEEE International Conference on Computer Vision. Venice, Italy, 2017:2980-2988.
[10] 刘俊明, 孟卫华. 基于深度学习的单阶段目标检测算法研究综述[J]. 航空兵器, 2020, 27(3):44-53.
LIU Junming, MENG Weihua. Review of single-stage target detection algorithm based on deep learning[J]. AERO weapons, 2020, 27(3):44-53.
[11] LIN T Y, DOLLáR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, USA, 2017:2117-2125.
[12] 陈景明, 金杰, 王伟锋. 基于特征金字塔网络的改进算法[J]. 激光与光电子学进展, 2019, 56(21):165-170
CHEN Jingming, JIN Jie, WANG Weifeng. Improved algorithm based on characteristic pyramid network[J]. Laser and optoelectronics progress, 2019, 56(21):165-170
[13] LAW H, DENG Jia. CornerNet:detecting objects as paired keypoints[C]//Proceedings of the 15th European Conference on Computer Vision (ECCV). Munich, Germany, 2018:734-750.
[14] NEWELL A, YANG Kaiyu, DENG Jia. Stacked hourglass networks for human pose estimation[C]//Proceedings of the 14th European Conference on Computer Vision. Amsterdam, The Netherlands, 2016:483-499.
[15] 和超, 张印辉, 何自芬. 多尺度特征融合工件目标语义分割[J]. 中国图象图形学报, 2020, 25(3):476-485
HE Chao, ZHANG Yinhui, HE Zifen. Multi scale feature fusion for object semantic segmentation[J]. Journal of image and graphic, 2020, 25(3):476-485
[16] CAO Guimei, XIE Xuemei, YANG Wenzhe, et al. Feature-fused SSD:fast detection for small objects[C]//Proceedings of SPIE 10615, Ninth International Conference on Graphic and Image Processing. Qingdao, China, 2018:106151E.
[17] 鞠默然, 罗江宁, 王仲博, 等. 一种融合注意力机制的多尺度目标检测算法[J]. 光学学报, 2020, 40(13):126-134.
JU Moran, LUO Jiangning, WANG Zhongbo, et al. Multi-scale target detection algorithm based on attention mechanism[J]. Acta optica sinica, 2020, 40(13):126-134.
[18] 王慧玲, 綦小龙, 武港山. 基于深度卷积神经网络的目标检测技术的研究进展[J]. 计算机科学, 2018, 45(9):11-19
WANG Huiling, QI Xiaolong, WU Gangshan. Research progress of target detection technology based on deep convolution neural network[J]. Computer science, 2018, 45(9):11-19
[19] TIAN Zhi, SHEN Chunhua, CHEN Hao, et al. FCOS:fully convolutional one-stage object detection[C]//Proceedings of 2019 IEEE/CVF International Conference on Computer Vision. Seoul, Korea, 2019:9627-9636.
[20] 李晓光, 付陈平, 李晓莉, 等. 面向多尺度目标检测的改进Faster R-CNN算法[J]. 计算机辅助设计与图形学学报, 2019, 31(7):1095-1101
LI Xiaoguang, FU Chenping, LI Xiaoli, et al. Improved faster R-CNN for multi-scale object detection[J]. Journal of computer-aided design and computer graphics, 2019, 31(7):1095-1101
[21] LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO:common objects in context[C]//Proceedings of the 13th European Conference on Computer Vision. Zurich, Switzerland, 2014:740-755.
[22] 李航, 朱明. 基于深度卷积神经网络的小目标检测算法[J]. 计算机工程与科学, 2020, 42(4):649-657
LI Hang, ZHU Ming. A small object detection algorithm based on deep convolutional neural network[J]. Computer engineering and science, 2020, 42(4):649-657

相似文献/References:

[1]赵文清,周震东,翟永杰.基于反卷积和特征融合的SSD小目标检测算法[J].智能系统学报,2020,15(2):310.[doi:10.11992/tis.201905035]
 ZHAO Wenqing,ZHOU Zhendong,ZHAI Yongjie.SSD small target detection algorithm based on deconvolution and feature fusion[J].CAAI Transactions on Intelligent Systems,2020,15(1):310.[doi:10.11992/tis.201905035]
[2]毛莺池,唐江红,王静,等.基于Faster R-CNN的多任务增强裂缝图像检测方法[J].智能系统学报,2021,16(2):286.[doi:10.11992/tis.201910004]
 MAO Yingchi,TANG Jianghong,WANG Jing,et al.Multi-task enhanced dam crack image detection based on Faster R-CNN[J].CAAI Transactions on Intelligent Systems,2021,16(1):286.[doi:10.11992/tis.201910004]
[3]齐鹏宇,王洪元,张继,等.基于改进FCOS的拥挤行人检测算法[J].智能系统学报,2021,16(4):811.[doi:10.11992/tis.202010012]
 QI Pengyu,WANG Hongyuan,ZHANG Ji,et al.Crowded pedestrian detection algorithm based on improved FCOS[J].CAAI Transactions on Intelligent Systems,2021,16(1):811.[doi:10.11992/tis.202010012]

备注/Memo

备注/Memo:
收稿日期:2020-04-27。
基金项目:国家自然科学基金项目(61871182);中央高校基本科研业务费面上项目(2020MS153)
作者简介:赵文清,教授,主要研究方向为人工智能与图像处理,主持或参与国家自然科学基金、河北省自然科学基金以及省部级项目10余项,获河北省科技进步二等奖1项、河北省科技进步三等奖1项。发表学术论文30余篇,出版学术专著1部;孔子旭,硕士研究生,主要研究方向为深度学习和目标检测;赵振兵,副教授,主要研究方向为深度学习与计算机视觉,主持或参与国家自然科学基金、河北省自然科学基金、北京市自然科学基金以及省部级项目10余项,获河北省科技进步一等奖1项。发表学术论文20余篇,出版学术专著3部
通讯作者:赵文清. E-mail:jbzwq@126.com
更新日期/Last Update: 2021-02-25