针对分布式光伏电站周界入侵检测数据不足、精度低、检测异物不完全等问题,在双模态特征融合基础上提出一种改进式的周界入侵目标检测算法（VT-DETR）。首先,将单通道网络拓展为双通道网络,以同时提取可见光与红外图像的多维特征;其次,引入软权重分配策略,以弥补可见光与红外图像融合时热辐射信息的缺失;最后,改善损失函数的计算方式,提升注意力机制训练效率。另外,构建面向分布式光伏电站的周界入侵图像数据集。实验结果表明,改进后算法的平均检测精度达到92.08%,验证了该方法在分布式光伏电站周界入侵检测上的有效性。
Abstract
An improved perimeter intrusion detection algorithm （visual thermal detection transformer, VT-DETR） is proposed based on bimodal feature fusion to solve the problems of insufficient perimeter intrusion detection data, low accuracy and incomplete detection of foreign objects in distributed photovoltaic power plants. Initially, the single-channel network is expanded into a dual-channel network, enabling simultaneous extraction of multidimensional features from visible light and infrared images. Subsequently, a strategy for soft weight allocation is introduced to compensate for the absence of thermal radiation information during the fusion of visible light and infrared images. Finally, the loss function calculation method is enhanced to improve the training efficiency of the attention mechanism. Additionally, a dedicated perimeter intrusion image dataset is built for the context of distributed photovoltaic power plants. Experimental results show the efficiency of the enhanced algorithm, achieving an average detection accuracy of 92.08% and thus validating its effectiveness to perimeter intrusion detection in distributed photovoltaic power plants.
关键词
深度学习 /
目标检测 /
图像融合 /
分布式光伏电站 /
周界入侵
Key words
deep learning /
object detection /
image fusion /
distributed photovoltaic power station /
perimeter intrusion
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 蒋兴群, 刘波, 宋力, 等. 基于改进YOLO-v3的风力机叶片表面损伤检测识别[J]. 太阳能学报, 2023, 44(3): 212-217.
JIANG X Q, LIU B, SONG L, et al.Surface damage detection and recognition of wind turbine blade based on improved YOLO-v3[J]. Acta energiae solaris sinica, 2023, 44(3): 212-217.
[2] 周颖, 王如意, 袁梓桐, 等. 一种高效双路径注意力太阳电池缺陷检测网络[J]. 太阳能学报, 2023, 44(4): 407-413.
ZHOU Y, WANG R Y, YUAN Z T, et al.An efficient dual-path attention solar cell defect detection network[J]. Acta energiae solaris sinica, 2023, 44(4): 407-413.
[3] 郭保青, 杨柳旭, 史红梅, 等. 基于快速背景差分的高速铁路异物侵入检测算法[J]. 仪器仪表学报, 2016, 37(6): 1371-1378.
GUO B Q, YANG L X, SHI H M, et al.High-speed railway clearance intrusion detection algorithm with fast background subtraction[J]. Chinese journal of scientific instrument, 2016, 37(6): 1371-1378.
[4] 管岭, 贾利民, 谢征宇. 融合注意力机制的轨道入侵异物检测轻量级模型研究[J]. 铁道学报, 2023, 45(5): 72-81.
GUAN L, JIA L M, XIE Z Y.Research on lightweight model for railway intrusion detection integrating attention mechanism[J]. Journal of the China Railway Society, 2023, 45(5): 72-81.
[5] 王瑞峰, 陈小屹. 基于改进YOLOv5的轨道异物入侵检测算法研究[J]. 云南大学学报(自然科学版), 2023, 45(4): 799-806.
WANG R F, CHEN X Y.Research on orbital foreign object intrusion detection algorithm based on improved YOLOv5[J]. Journal of Yunnan University (natural sciences edition), 2023, 45(4): 799-806.
[6] 肖曾翔, 徐启峰. 基于改进卷积神经网络的变电站异物入侵识别[J]. 科学技术与工程, 2022, 22(4): 1465-1471.
XIAO Z X, XU Q F.Recognition of foreign objects intrusion in substation based on improved convolutional neural network[J]. Science technology and engineering, 2022, 22(4): 1465-1471.
[7] 宁欣, 田伟娟, 于丽娜, 等. 面向小目标和遮挡目标检测的脑启发CIRA-DETR全推理方法[J]. 计算机学报, 2022, 45(10): 2080-2092.
NING X, TIAN W J, YU L N, et al.Brain-inspired CIRA-DETR full inference model for small and occluded object detection[J]. Chinese journal of computers, 2022, 45(10): 2080-2092.
[8] YANG L, MA R, ZAKHOR A. Drone object detection using rgb/ir fusion[J]. arXiv preprint arXiv :2201.03786, 2022.
[9] DEVAGUPTAPU C, AKOLEKAR N, SHARMA M M, et al.Borrow from anywhere: pseudo multi-modal object detection in thermal imagery[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Long Beach, CA, USA, 2019: 1029-1038.
[10] GAO C Q, MENG D Y, YANG Y, et al.Infrared patch-image model for small target detection in a single image[J]. IEEE transactions on image processing, 2013, 22(12): 4996-5009.
[11] LI Y M, ZHANG W, LIU Y Y, et al.An efficient fire and smoke detection algorithm based on an end-to-end structured network[J]. Engineering applications of artificial intelligence, 2022, 116: 105492.
[12] CARION N, MASSA F, SYNNAEVE G, et al.End-to-end object detection with transformers[M]. Cham: Springer International Publishing, 2020: 213-229.
[13] 徐硕, 邓艾东, 杨宏强, 等. 基于改进残差网络的旋转机械故障诊断[J]. 太阳能学报, 2023, 44(7): 409-418.
XU S, DENG A D, YANG H Q, et al.Rotating machinery fault diagnosis method based on improved residual neural network[J]. Acta energiae solaris sinica, 2023, 44(7): 409-418.
[14] LEYKIN A, RAN Y, HAMMOUD R.Thermal-visible video fusion for moving target tracking and pedestrian classification[C]//2007 IEEE Conference on Computer Vision and Pattern Recognition, Minneapolis, MN, USA, 2007: 1-8.
[15] 赵明, 张浩然. 一种基于跨域融合网络的红外目标检测方法[J]. 光子学报, 2021, 50(11): 339-349.
ZHAO M, ZHANG H R.An infrared object detection method based on cross-domain fusion network[J]. Acta photonica sinica, 2021, 50(11): 339-349.
[16] SHAW P, USZKOREIT J, VASWANI A. Self-attention with relative position representations[J]. arXiv preprint arXiv:1803. 02155,2018.
[17] DAI Y N, YU J Y, ZHANG D A, et al.RODFormer: high-precision design for rotating object detection with transformers[J]. Sensors, 2022, 22(7): 2633.
基金
浙江省重点研发计划（2021C04030）;浙江省基础公益研究计划（LGG21F030004）;浙江省“尖兵”“领雁”研发攻关计划（2022C04012）
PDF(2983 KB)
