Abstract
Aiming at the problem that the target pedestrians’ environment in the intelligent driving system is complex and the detection accuracy is not high due to large size changes, the YOLOv3 algorithm is improved in this paper and the K-means clustering method to directly calculate the anchor frame size of the data set is used. Perform cluster analysis to reduce network training time.The multi-scale fusion of shallow features and deep features improves the feature extraction effect of the skeleton network; the SE module is introduced to enhance the spatial channel after the multi-level feature fusion, which promotes the ability of pedestrian multi-scale information mining, and integrates the DIoU loss function Introducing the training process speeds up the model convergence and running speed.The model was trained with VOC2007, VOC2012 and Caltech datasets, and the Caltech pedestrian dataset was used to verify the effectiveness of the improved algorithm. The results show that the improved YOLOv3 algorithm for small-scale pedestrian targets in the intelligent driving system has improved detection accuracy.
This paper is supported by the Project of Jilin Province Science and Technology Department Plan (No. 20200703016ZP), the Projects in Science and Technology of the 13th Five-year Plan of Education Department of Jilin Province (No. JJKH20200671KJ)
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Krizhevsky, A., Sutskever, I., Hinton, G.E.: ImageNet classification with deep convolutional neural networks. In: Proceedings of the 25th International Conference on Neural Information ProcessingSystems.Lake Tahoe, Nevada, USA, pp. 1097–1105 (2012)
Wen, Z., Du, Y.H., Yoshida, T., et al.: DRI-RCNN: an approach to deceptive review identification using recurrent convolutional neural network. Inf. Process. Manag. 54(4), 576–592 (2018)
Girshick, R.: Fast R-CNN. In: 2015 IEEE International Conference on Computer Vision (ICCV), Santiago, pp. 1440–1448 (2015)
Ren, S., He, K., Girshick, R., et al.: Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Trans. Pattern Anal. Mach. Intell. 39(6), 1137–1149 (2017)
Liu, W., et al.: Ssd: single shot multibox detector. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) Computer Vision – ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I, pp. 21–37. Springer International Publishing, Cham (2016). https://doi.org/10.1007/978-3-319-46448-0_2
Fu, C.Y., Liu, W., Ranga, A., et al.: DSSD: deconvolutional single shot detector. In: European Conference on Computer Vision and Pattern Recognition, pp. 910–925 (2017)
Li, Z., Zhou, F.: FSSD: feature fusion single shot multi-box detector. arXiv: 1512.02325 (2017)
Jisoo, J., Hyojin, P., Nojun, K.: Enhancement of SSD by concatenating feature maps for object detection. arXiv: 1705.09587 (2017)
Redmon, J., Divvala, S., Girshick, R., et al.: You only look once: unified, real time object detection. In: Proceedings of Computer Vision and Pattern Recognition, pp. 779–788. IEEE Press, Washington D.C. (2015)
Redmon, J., Farhadi, A.: YOLO9000: better, faster, stronger. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 7263–7271 (2017)
Redmon, J.: YOLOv3: an incremental improvement (2018)
Zhang, Y., Shen, Y.L., Zhang, J.: An improved Tiny-YOLOv3 pedestrian detection algorithm. Optik 183, 17–23 (2019)
Wang, G.W.: Multi-Scale Feature Fusion Improves Pedestrian and Vehicle Detection in YOLOv3 network. Nanjing University of Posts and Telecommunications, Nanjing (2019)
Shi, Z., Mao, L., Sun, J.: Multi-modal weighted fusion pedestrian detection algorithm based on YOLO. Comput. Eng., 1 (2021)
He, K.M., Zhang, X.Y., Ren, S.Q., et al.: Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778 (2016)
Hu, J., Li, S., Albanie, S., et al.: Squeeze and excitation network (2018)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Liu, S., Geng, Y., Song, Y., Yan, W., Lian, Y. (2022). Research on Small Target Pedestrian Detection Algorithm Based on Improved YOLOv3. In: Chu, SC., Lin, J.CW., Li, J., Pan, JS. (eds) Genetic and Evolutionary Computing. ICGEC 2021. Lecture Notes in Electrical Engineering, vol 833. Springer, Singapore. https://doi.org/10.1007/978-981-16-8430-2_19
Download citation
DOI: https://doi.org/10.1007/978-981-16-8430-2_19
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-16-8429-6
Online ISBN: 978-981-16-8430-2
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)