面向路侧交通监控场景的轻量车辆检测模型

doi:10.3778/j.issn.1002-8331.2109-0516

摘要/Abstract

摘要： 针对路侧交通监控场景和智能交通管控需要，提出轻量型的车辆检测算法GS-YOLO，解决现有模型检测速度慢、占用内存多的问题。GS-YOLO借鉴GhostNet思想将传统卷积分为两步，利用轻量操作增强特征，降低模型的计算量。在主干特征提取网络中引入注意力机制，对重要信息进行选择，提高模块的检测能力。另外参考SqueezeNet结构，使用Fire Module和深度可分离卷积减少模型参数，模型大小从244?MB降低到34?MB，内存占用降低了86%。使用Roofline模型对实验数据和模型实际性能进行分析，结果表明GS-YOLO的精确度（AP）达到85.55%，相比YOLOv4提升了约0.45%。由于受计算平台带宽影响，GS-YOLO在GPU上检测速度提升7.3%，但在CPU上检测速度提高了83%，更适用于算力资源不足的小型设备。

关键词: 图像处理, 目标检测, 轻量化, GhostNet, 深度可分离卷积

Abstract: To satisfy the needs of roadside traffic monitoring scenarios and intelligent traffic control, a lightweight vehicle detection model GS-YOLO is proposed to solve the problems of low detection efficiency and high memory consumption of existing models. Following the structure of GhostNet, GS-YOLO divides the vanilla convolution into two steps and uses cheap operations to enhance features and reduces the resource consumption of the model. An attention mechanism is introduced in the backbone feature extraction network to select important information and improve the detection capability of GS-YOLO. The model parameters are reduced by using Fire Module and depthwise separable convolution with reference to the SqueezeNet, and the model size is reduced from 244 MB to 34 MB, with an 86% reduction in memory usage. Using Roofline theory to analyze the experimental data and the actual performance of the model. The experimental results show that the accuracy（AP） of GS-YOLO reaches 85.55%, which is about 0.45% higher than that of YOLOv4. Due to the bandwidth impact of computing platform, GS-YOLO has a 7.3% improvement in image processing speed on GPU and 83% on CPU, which is more suitable for small devices with insufficient computing power resources.

Key words: image processing, object detection, lightweight, GhostNet, depthwise separable convolution

郭宇阳, 胡伟超, 戴帅, 陈艳艳. 面向路侧交通监控场景的轻量车辆检测模型[J]. 计算机工程与应用, 2022, 58(6): 192-199.

GUO Yuyang, HU Weichao, DAI Shuai, CHEN Yanyan. Lightweight Vehicle Detection Model for Roadside Traffic Monitoring Scenarios[J]. Computer Engineering and Applications, 2022, 58(6): 192-199.

参考文献

[1] LIU W，ANGUELOV D，ERHAN D，et al.SSD：single shot multibox detector[C]//European Conference on Computer Vision，Amsterdam，Oct 11-14，2016.Berlin：Springer，2016：21-37.
[2] LIN T Y，GOYAL P，GIRSHICK R，et al.Focal loss for dense object detection[C]//Proceedings of the IEEE International Conference on Computer Vision，Venice，Oct 22-29，2017.Washington D C：IEEE，2017：2980-2988.
[3] 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，Las Vegas，Jun 26-Jul 1，2016.Piscataway：IEEE，2016：779-788.
[4] REDMON J，FARHADI A.YOLO9000：better，faster，stronger[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，Hawaii，Jul 21-26，2017.Piscataway：IEEE，2017：7263-7271.
[5] FARHADI A，REDMON J.Yolov3：an incremental improvement[C]//Computer Vision and Pattern Recognition，Salt Lake City，Jun 18-22，2018.Piscataway：IEEE，2018.
[6] REN S，HE K，GIRSHICK R，et al.Faster R-CNN：towards real-time object detection with region proposal networks[J].Advances in Neural Information Processing Systems，2015，28：91-99.
[7] CAI Z，VASCONCELOS N.Cascade R-CNN：delving into high quality object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，Salt Lake City，Jun 18-22，2018.Piscataway：IEEE，2018：6154-6162.
[8] LAW H，DENG J.Cornernet：detecting objects as paired keypoints[C]//Proceedings of the European Conference on Computer Vision，Munich，Sept 8-14，2018.Berlin：Springer，2018：734-750.
[9] DUAN K，BAI S，XIE L，et al.Centernet：keypoint triplets for object detection[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision Long Beach，Jun 16-20，2019.Piscataway：IEEE，2019：6569-6578.
[10] 曹诗雨，刘跃虎，李辛昭.基于Fast R-CNN的车辆目标检测[J].中国图象图形学报，2017，22（5）：671-677.
CAO S Y，LIU Y H，LI X Z.Vehicle detection method based on fast R-CNN[J].Journal of Image and Graphics，2017，22（5）：671-677.
[11] 李松江，吴宁，王鹏，等.基于改进Cascade RCNN的车辆目标检测方法[J].计算机工程与应用，2021，57（5）：123-130.
LI S J，WU N，WANG P，et al.Vehicle target detection method based on improved Cascade RCNN[J].Computer Engineering and Applications，2021，57（5）：123-130.
[12] 杜金航，何宁.基于改进的YOLOv3道路车辆实时检测[J].计算机工程与应用，2020，56（11）：26-32.
DU J H，HE N.Real-time road vehicles detection based on improved YOLOv3[J].Computer Engineering and Applications，2020，56（11）：26-32.
[13] 金旺，易国洪，洪汉玉，等.基于卷积神经网络的实时车辆检测[J].计算机工程与应用，2021，57（5）：222-228.
JIN W，YI G H，HONG H Y，et al.Real-time vehicle detection based on convolutional neural network[J].Computer Engineering and Applications，2021，57（5）：222-228.
[14] 顾恭，徐旭东.改进YOLOv3的车辆实时检测与信息识别技术[J].计算机工程与应用，2020，56（22）：173-184.
GU G，XU X D.Real-time vehicle detection and information recognition technology based on YOLOv3 improved algorithm[J].Computer Engineering and Applications，2020，56（22）：173-184.
[15] 鲁博，瞿绍军.融合BiFPN和改进Yolov3-tiny网络的航拍图像车辆检测方法[J].小型微型计算机系统，2021，42（8）：1694-1698.
LU B，ZHAI S J.Vehicle detection method in aerial images based on BiFPN and improved Yolov3-tiny network[J].Journal of Chinese Computer Systems，2021，42（8）：1694-1698.
[16] 李宇昕，杨帆，刘钊等.基于改进残差网络的道口车辆分类方法[J].激光与光电子学进展，2021，58（4）：384-390.
LI Y X，YANG F，LIU Z，et al.Classification method of crossing vehicle based on improved residual network[J].Laser & Optoelectronics Progress，2021，58（4）：384-390.
[17] 李汉冰，徐春阳，胡超超.基于YOLOV3改进的实时车辆检测方法[J].激光与光电子学进展，2020，57（10）：332-338.
LI H B，XU C Y，HU C C.Improved real-time vehicle detection method based on YOLOV3[J].Laser & Optoelectronics Progress，2020，57（10）：332-338.
[18] GEIGER A，LENZ P，STILLER C，et al.Vision meets robotics：the kitti dataset[J].The International Journal of Robotics Research，2013，32（11）：1231-1237.
[19] WEN L，DU D，CAI Z，et al.UA-DETRAC：a new benchmark and protocol for multi-object detection and tracking[J].Computer Vision and Image Understanding，2020，193：102907.
[20] BOCHKOVSKIY A，WANG C Y，LIAO H Y M.Yolov4：optimal speed and accuracy of object detection[EB/OL].（2020-04-23）[2021-07-05].http：//arxiv.org/abs/2004.10934.
[21] IANDOLA F N，HAN S，MOSKEWICZ M W，et al.SqueezeNet：AlexNet-level accuracy with 50x fewer parameters and<0.5MB model size[EB/OL].（2016-11-04）[2021-08-21].http：//arxiv.org/abs/1602.07360.
[22] HAN K，WANG Y，TIAN Q，et al.Ghostnet：more features from cheap operations[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition，Washington D C，Jun 14-19，2020.Piscataway：IEEE，2020：1580-1589.
[23] WILLIAMS S，WATERMAN A，PATTERSON D.Roofline：an insightful visual performance model for multicore architectures[J].Communications of the ACM，2009，52（4）：65-76.