Abstract
Object detection is one of the widest studies in computer vision. And there are many useful algorithms for this task. Though most of the studies are effective in the daytime or dark scene, they are not applicable in some special cases that need to carry out object detection in the scenarios for both daytime as well as night such as rescue scenarios in serious disasters of earthquake, landslide, mountain torrent, flood and on. To solve this issue, we attend to study the object detection algorithm in the scenarios for both daytime as well as dark scene. Considering the high efficiency of the YOLO algorithm under sufficient light conditions as well as the needs of object detection under poor light, this paper proposes an improved end-to-end YOLOv3 network under dark conditions. The main idea is to integrate the YOLOv3 object detection network and the Retinex image enhancement network to ensure that the image be easier to perform feature extraction and object recognition. Extensive experiments were conducted on several public datasets. The results show that compared with the traditional YOLOv3 object detection model, the proposed algorithm can improve the mean average precision from 53.23% to 59.79% in the object detection task with poor light.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Chen, C., Chen, Q., Xu, J., Koltun, V.: Learning to see in the dark. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 3291–3300 (2018)
Dong, X., et al.: Fast efficient algorithm for enhancement of low lighting video. In: 2011 IEEE International Conference on Multimedia and Expo, pp. 1–6. IEEE (2011)
Everingham, M., Eslami, S., Gool, L.V., Williams, C., Winn, J., Zisserman, A.: The pascal visual object classes challenge: a retrospective. Int. J. Comput. Vis. 111(1), 98–136 (2015)
Girshick, R.: Fast R-CNN. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 1440–1448 (2015)
Jobson, D.J., Rahman, Z.U., Woodell, G.A.: A multiscale retinex for bridging the gap between color images and the human observation of scenes. IEEE Trans. Image Process. 6(7), 965–976 (1997)
Li, Z., Song, X., Chen, C., Wang, C.: Brightness level image enhancement algorithm based on retinex algorithm. J. Data Acquisit. Process. 41–49 (2019)
Liu, Z., Lu, Y., Tang, X., Uyttendaele, M., Jian, S.: Fast burst images denoising. ACM Trans. Graph. 33(6CD), 232.1-232.9 (2014)
Pu, T., Zhang, Z., Peng, Z.: Enhancing uneven lighting images with naturalness preserved retinex algorithm. J. Data Acquisit. Process. 36, 76–84 (2021)
Redmon, J., Divvala, S., Girshick, R., Farhadi, A.: You only look once: unified, real-time object detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 779–788 (2016)
Redmon, J., Farhadi, A.: Yolo9000: better, faster, stronger. In: IEEE Conference on Computer Vision & Pattern Recognition, pp. 6517–6525 (2017)
Redmon, J., Farhadi, A.: Yolov3: an incremental improvement. CoRR abs/1804.02767 (2018), http://arxiv.org/abs/1804.02767
Ren, S., He, K., Girshick, R., Sun, J.: Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Trans. Pattern Anal. Mach. Intell. 39(6), 1137–1149 (2017)
Sasagawa, Y., Nagahara, H.: YOLO in the dark - domain adaptation method for merging multiple models. In: Vedaldi, A., Bischof, H., Brox, T., Frahm, J.-M. (eds.) ECCV 2020. LNCS, vol. 12366, pp. 345–359. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58589-1_21
Wang, J., Chen, K., Yang, S., Loy, C.C., Lin, D.: Region proposal by guided anchoring. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 2965–2974 (2019)
Wei, C., Wang, W., Yang, W., Liu, J.: Deep retinex decomposition for low-light enhancement. arXiv preprint arXiv:1808.04560 (2018)
Xu, J., Dou, Y., Zheng, Y.: Underwater target recognition and tracking method based on YOLO-V3 algorithm. J. Chin. Inertial Technol. 28, 129–133 (2020)
Acknowledgements
As an undergraduate, it is the first time for me to present my results in KSEM2021. I cannot express my excitement. First of all, I’d like to thank my supervisor, Professor Peng Junjie. He is knowledgeable and rigorous in his scholarship. He gave me good suggestions on the general direction of my thesis and carefully guided my thesis writing. At the same time, I also want to thank my senior male Yuan Haochen and senior female Wu Ting, who gave me a lot of help in revising the paper. Finally, I would like to express my heartfelt thanks to my teachers and the people who care and support me.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Wang, J. (2021). An Improved YOLO Algorithm for Object Detection in All Day Scenarios. In: Qiu, H., Zhang, C., Fei, Z., Qiu, M., Kung, SY. (eds) Knowledge Science, Engineering and Management. KSEM 2021. Lecture Notes in Computer Science(), vol 12817. Springer, Cham. https://doi.org/10.1007/978-3-030-82153-1_39
Download citation
DOI: https://doi.org/10.1007/978-3-030-82153-1_39
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-82152-4
Online ISBN: 978-3-030-82153-1
eBook Packages: Computer ScienceComputer Science (R0)