Logo PTI Logo FedCSIS

Communication Papers of the 18th Conference on Computer Science and Intelligence Systems

Annals of Computer Science and Information Systems, Volume 37

Real-Time Detection of Small Objects in Automotive Thermal Images with Modern Deep Neural Architectures

, ,

DOI: http://dx.doi.org/10.15439/2023F8409

Citation: Communication Papers of the 18th Conference on Computer Science and Intelligence Systems, M. Ganzha, L. Maciaszek, M. Paprzycki, D. Ślęzak (eds). ACSIS, Vol. 37, pages 2935 ()

Full text

Abstract. Thermal imaging has shown great potential for improving object detection in automotive settings, particularly in low light or adverse weather conditions. To help and further develop this industry, we extend our previously shared Thermal Automotive Dataset by more than 2000 new images and 2 novel object detecting models based on YOLOv5 and YOLOv7 architecture. We point how important is the size of the dataset. Additionally, we compare the performance of both models, to see which is more reliable and superior in terms of detecting small objects in thermal spectrum. Furthermore, we analysed how preprocessing affects thermal imaging dataset and models basing on it. The new dataset is available free from the Internet.


  1. M. Knapik and B. Cyganek, “Fast eyes detection in thermal images,” Multimedia Tools and Applications, vol. 80, no. 3, pp. 3601–3621, 2021.
  2. ——, “Driver’s fatigue recognition based on yawn detection in thermal images,” Neurocomputing, vol. 338, pp. 274–292, 2019. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0925231219302280
  3. M. A. Farooq, W. Shariff, D. O’callaghan, A. Merla, and P. Corcoran, “On the role of thermal imaging in automotive applications: A critical review,” IEEE Access, vol. 11, pp. 25 152–25 173, 2023.
  4. T. Balon, M. Knapik, and B. Cyganek, “New thermal automotive dataset for object detection,” Annals of Computer Science and Information Systems, vol. 31, pp. 43–48, 2022.
  5. G. Jocher, A. Chaurasia, A. Stoken, J. Borovec, NanoCode012, Y. Kwon, TaoXie, J. Fang, imyhxy, K. Michael, Lorna, A. V, D. Montes, J. Nadar, Laughing, tkianai, yxNONG, P. Skalski, Z. Wang, A. Hogan, C. Fati, L. Mammana, AlexWang1900, D. Patel, D. Yiwei, F. You, J. Hajek, L. Diaconu, and M. T. Minh, “ultralytics/yolov5: v6.1 - TensorRT, TensorFlow Edge TPU and OpenVINO Export and Inference,” Feb. 2022. [Online]. Available: https://doi.org/10.5281/zenodo.6222936
  6. C.-Y. Wang, A. Bochkovskiy, and H.-Y. M. Liao, “Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors,” 2022.
  7. “Automotivethermo2_0,” 2023. [Online]. Available: https://home.agh.edu.pl/~cyganek/AutomotiveThermo2_0.zip
  8. A. Bochkovskiy, C.-Y. Wang, and H.-Y. M. Liao, “Yolov4: Optimal speed and accuracy of object detection,” 2020.
  9. C.-Y. Wang, H.-Y. M. Liao, I.-H. Yeh, Y.-H. Wu, P.-Y. Chen, and J.-W. Hsieh, “Cspnet: A new backbone that can enhance learning capability of cnn,” 2019.
  10. G. Huang, Z. Liu, L. van der Maaten, and K. Q. Weinberger, “Densely connected convolutional networks,” 2018.
  11. M. Tan and Q. V. Le, “Efficientnet: Rethinking model scaling for convolutional neural networks,” 2020.
  12. C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, and A. Rabinovich, “Going deeper with convolutions,” 2014.
  13. “Darklabel annotation software,” https://github.com/darkpgmr/DarkLabel, accessed: 2022-06-07.
  14. “Sudden performance decrease in training · Issue 5721 · ultralyt-ics/yolov5 — github.com,” https://github.com/ultralytics/yolov5/issues/ 5721, 2021, [Accessed 31-07-2023].
  15. “Unstable training · Issue 974 · WongKinYiu/yolov7 — github.com,” https://github.com/WongKinYiu/yolov7/issues/974, 2022, [Accessed 31-07-2023].
  16. B. Cyganek and M. Woźniak, “Virtual high dynamic range imaging for underwater drone navigation,” Proceedings of the 6th IIAE International Conference on Industrial Application Engineering, pp. 393–398, 2018.
  17. A. Song, Z. Zhao, Q. Xiong, and J. Guo, “Lightweight the focus module in yolov5 by dilated convolution,” in 2022 3rd International Conference on Computer Vision, Image and Deep Learning International Conference on Computer Engineering and Applications (CVIDL ICCEA), 2022, pp. 111–114.
  18. L. Yang, “Investigation of you only look once networks for vision-based small object detection,” International Journal of Advanced Computer Science and Applications, vol. 14, no. 4, 2023. [Online]. Available: http://dx.doi.org/10.14569/IJACSA.2023.0140410
  19. O. E. Olorunshola, M. E. Irhebhude, and A. E. Evwiekpaefe, “A comparative study of yolov5 and yolov7 object detection algorithms,” Journal of Computing and Social Informatics, vol. 2, no. 1, pp. 43–48, 2023.