Fusion of Depth and Thermal Imaging for People Detection
DOI:
https://doi.org/10.26636/jtit.2021.155521Keywords:
depth imaging, person detection, sensors fusion, thermal imagingAbstract
The methodology presented in this paper covers the topic of automatic detection of humans based on two types of images that do not rely on the visible light spectrum, namely on thermal and depth images. Various scenarios are considered with the use of deep neural networks being extensions of Faster R-CNN models. Apart from detecting people, independently, with the use of depth and thermal images, we proposed two data fusion methods. The first approach is the early fusion method with a 2-channel compound input. As it turned out, its performance surpassed that of all other methods tested. However, this approach requires that the model be trained on a dataset containing both types of spatially and temporally synchronized imaging sources. If such a training environment cannot be setup or if the specified dataset is not sufficiently large, we recommend the late fusion scenario, i.e. the other approach explored in this paper. Late fusion models can be trained with single-source data. We introduce the dual-NMS method for fusing the depth and thermal imaging approaches, as its results are better than those achieved by the common NMS.
Downloads
References
[1] A. Clapés, J. S. Jacques Junior, C. Morral, and S. Escalera, „Chalearn lap 2020 challenge on identity-preserved human detection: Dataset and results", in 15th IEEE Int. Conf. on Automatic Face and Gesture Recogn. (FG 2020), Buenos Aires, Argentina, pp. 859-866, 2020. DOI: https://doi.org/10.1109/FG47880.2020.00135
View in Google Scholar
[2] T.-Y. Lin et al., „Microsoft COCO: common objects in context", CoRR, abs/1405.0312, 2014 [Online]. Available: https://arxiv.org/pdf/1405.0312
View in Google Scholar
[3] Teledyne Flir LLC, „LWIR micro thermal camera module Lepton 3", 2018 [Online]. Available: https://www.ir.com/products/lepton/?model=500-0276-01 (accessed on: 01.01.2021).
View in Google Scholar
[4] J. Smisek, M. Jancosek, and T. Pajdla, 3D with Kinect, Consumer Depth Cameras for Computer Vision. Adv. in Computer Vision and Pattern Recogn., A. Fossati, J. Gall, H. Grabner, X. Ren, K. Konolige, Eds., pp. 3-25. Springer London, London: 2013. DOI: https://doi.org/10.1007/978-1-4471-4640-7_1
View in Google Scholar
[5] Intel Corporation, Intel RealSense Depth Module D400 Series CustomCalibration, 2019 [Online]. Available: https://www.intel.com/content/dam/support/ us/en/documents/emerging-technologies/ intel-realsense-technology/RealSense D400% 20 Custom Calib Paper.pdf (accessed on: 01.01.2021).
View in Google Scholar
[6] S. Kumar, T. K. Marks, and M. Jones, „Improving person tracking using an inexpensive thermal infrared sensor", in IEEE Conf. On Computer Vision and Pattern Recogn. Workshops, Columbus, OH, USA, pp. 217-224, 2014. DOI: https://doi.org/10.1109/CVPRW.2014.41
View in Google Scholar
[7] A. S. Charan, M. Jitesh, M. Chowdhury, and H. Venkataraman, „Abifn: Attention-based bi-modal fusion network for object detection at night time", Electronics Letters, vol. 56, no. 24, pp. 1309-1311, 2020. DOI: https://doi.org/10.1049/el.2020.1952
View in Google Scholar
[8] H. Haggag, M. Hossny, S. Nahavandi, and O. Haggag, „An adaptable system for RGB-D based human body detection and pose estimation: Incorporating attached props", in IEEE Int. Conf. on Systems, Man, and Cybernetics (SMC), Budapest, Hungary, pp. 1544-1549, 2016. DOI: https://doi.org/10.1109/SMC.2016.7844458
View in Google Scholar
[9] O. H. Jafari, D. Mitzel, and B. Leibe, „Real-time RGB-D based people detection and tracking for mobile robots and headworn cameras", in IEEE Int. Conf. on Robotics and Automation (ICRA), Hong Kong, China, 2014, pp. 5636-5643. DOI: https://doi.org/10.1109/ICRA.2014.6907688
View in Google Scholar
[10] M. Rasoulidanesh, S. Yadav, S. Herath, Y. Vaghei, and S. Payandeh, „Deep attention models for human tracking using RGBD", Sensors, vol. 19, no. 4, 2019. DOI: https://doi.org/10.3390/s19040750
View in Google Scholar
[11] H. S. Hadi, M. Rosbi, U. U. Sheikh, and S. H. M. Amin, „Fusion of thermal and depth images for occlusion handling for human detection from mobile robot", in 10th Asian Control Conf. (ASCC), Kota Kinabalu, Malaysia, pp. 1-5, 2015. DOI: https://doi.org/10.1109/ASCC.2015.7244722
View in Google Scholar
[12] D. J. Yeong, G. Velasco-Hernandez, J. Barry, and J. Walsh, „Sensor and sensor fusion technology in autonomous vehicles: A review", Sensors, vol. 21, no. 6, 2021. DOI: https://doi.org/10.3390/s21062140
View in Google Scholar
[13] V. F. Vidal et al., „Sensors fusion and multidimensional point cloud analysis for electrical power system inspection", Sensors, vol. 20, no. 14, pp. 40-42, 2020. DOI: https://doi.org/10.3390/s20144042
View in Google Scholar
[14] T. Alldieck, C. H. Bahnsen, and T. B. Moeslund, „Context-aware fusion of RGB and thermal imagery for traffic monitoring", Sensors, vol. 16, no. 11, 2016. DOI: https://doi.org/10.3390/s16111947
View in Google Scholar
[15] F. Farahnakian and J. Heikkonen, „Deep learning based multi-modal fusion architectures for maritime vessel detection", Remote Sensing, vol. 12, no. 16, 2020. DOI: https://doi.org/10.3390/rs12162509
View in Google Scholar
[16] A. Morfin-Santana et al., „Real-time people detection from thermal images by using an unmanned aerial system", in 16th Int. Conf. on Electric. Engineer., Comput. Sci. and Automatic Control (CCE), Mexico City, Mexico, pp. 1-6, 2019. DOI: https://doi.org/10.1109/ICEEE.2019.8884561
View in Google Scholar
[17] S. Chang, F. Yang, W. Wu, Y. Cho, and S. Chen, „Nighttime pedestrian detection using thermal imaging based on hog feature", in Proc. 2011 Int. Conf. on System Sci. and Engineer., Macau, China, pp. 694-698, 2011. DOI: https://doi.org/10.1109/ICSSE.2011.5961992
View in Google Scholar
[18] L. Spinello and K. O. Arras, „People detection in RGB-D data", in IEEE/RSJ Int. Conf. on Intell. Robots and Systems, San Francisco, CA, USA, pp. 3838-3843, 2011. DOI: https://doi.org/10.1109/IROS.2011.6095074
View in Google Scholar
[19] Ch. Herrmann, M. Ruf, and J. Beyerer, „CNN-based thermal infrared person detection by domain adaptation", in Proc. Autonomous Systems: Sensors, Vehicles, Security, and the Internet of Everything, M. C. Dudzik and J. C. Ricklin, Eds., Orlando, FL, USA, vol. 10643, 2018, pp. 38-43.
View in Google Scholar
[20] Sh. Ren, K. He, R. Girshick, and J. Sun, „Faster R-CNN: Towards real-time object detection with region proposal networks", in Proc. of the 28th Int. Conf. on Neural Information Process. Systems, C. Cortes, N. D. Lawrence, D. D. Lee, M. Sugiyama, and R. Garnett, Eds., vol. 1, pp. 91-99, 2015 [Online]. Available: http://papers.nips.cc/paper/5638-faster-r-cnn-towards-real-time-object-detection-with-region-proposal-networks.pdf
View in Google Scholar
[21] K. He, G. Gkioxari, P. Dollár, and R. B. Girshick, „Mask R-CNN", CoRR, abs/1703.06870, 2017 [Online]. Available: https://arxiv.org/pdf/1703.06870
View in Google Scholar
[22] W. Liu et al., „SSD: single shot multibox detector", CoRR, abs/1512.02325, 2015 [Online]. Available: https://arxiv.org/pdf/1512.02325
View in Google Scholar
[23] J. Redmon and A. Farhadi, „Yolov3: An incremental improvement", CoRR, abs/1804.02767, 2018 [Online]. Available: https://arxiv.org/pdf/1804.02767
View in Google Scholar
[24] K. He, X. Zhang, Sh. Ren, and J. Sun, „Deep residual learning for image recognition", CoRR, abs/1512.03385, 2015 [Online]. Available: https://arxiv.org/pdf/1512.03385
View in Google Scholar
[25] A. Karpathy et al., „Large-scale video classification with convolutional neural networks", 27th IEEE Conf. on Computer Vision and Pattern Recogn. (CVPR), Columbus, OH, USA, 2014. DOI: https://doi.org/10.1109/CVPR.2014.223
View in Google Scholar
[26] A. Neubeck and L. Van Gool, „Efficient non-maximum suppression", in 18th Int. Conf. on Pattern Recogn. (ICPR'06), Hong Kong, China, vol. 3, pp. 850-855, 2006. DOI: https://doi.org/10.1109/ICPR.2006.479
View in Google Scholar
Downloads
Submitted
Published
Issue
Section
License
Copyright (c) 2021 Journal of Telecommunications and Information Technology
This work is licensed under a Creative Commons Attribution 4.0 International License.
