A Vision-Based On-road Vehicle Light Detection System Using Support Vector Machines

  • J. ArunnehruEmail author
  • H. Anwar Basha
  • Ajay Kumar
  • R. Sathya
  • M. Kalaiselvi Geetha
Part of the Studies in Computational Intelligence book series (SCI, volume 771)


Vehicle light detection and recognition for collision avoidance presents a major challenge in urban driving conditions. In this chapter, an optical flow method is used to extract moving vehicles in a traffic environment, and hue-saturation-value (HSV) color space is adopted to detect vehicle brake and turn light indicators. In addition, a morphological operation is applied to obtain the precise vehicle light region. The proposed Vehicle Light Block Intensity Vector (VLBIV) feature extraction from the vehicle light region is realized by a supervised learning method known as support vector machines (SVM). Analysis is carried out on the vehicle signal recognition system which interprets the color videos taken from a front-view video camera of a car operating in traffic scenarios. This technique yields average accuracy of 98.83% in SVM (RBF) in 36 VLBIV features when compared to an SVM (polynomial) classifier.


Optical flow HSV color space Support vector machine Performance measure 


  1. 1.
    Arunnehru, J., and M. Kalaiselvi Geetha. 2013. Motion intensity code for action recognition in video using PCA and SVM. In Mining Intelligence and Knowledge Exploration, 70–81. Cham: Springer.CrossRefGoogle Scholar
  2. 2.
    Lu, Jianbo, Hassen Hammoud, Todd Clark, Otto Hofmann, Mohsen Lakehal-ayat, Shweta Farmer, Jason Shomsky, and Roland Schaefer. 2017. A System for Autonomous Braking of a Vehicle Following Collision. No. 2017–01–1581. SAE Technical Paper.Google Scholar
  3. 3.
    Zhang, Bailing, Yifan Zhou, Hao Pan, and Tammam Tillo. 2014. Hybrid model of clustering and kernel autoassociator for reliable vehicle type classification. Machine Vision and Applications 25 (2): 437–450.CrossRefGoogle Scholar
  4. 4.
    Almagambetov, Akhan, Senem Velipasalar, and Mauricio Casares. 2015. Robust and computationally lightweight autonomous tracking of vehicle taillights and signal detection by embedded smart cameras. IEEE Transactions on Industrial Electronics 62 (6): 3732–3741.CrossRefGoogle Scholar
  5. 5.
    Hillel, Aharon Bar, Ronen Lerner, Dan Levi, and Guy Raz. 2014. Recent progress in road and lane detection: a survey. Machine Vision And Applications 25 (3): 727–745.Google Scholar
  6. 6.
    Nourani-Vatani, Navid, Paulo Vinicius Koerich Borges, Jonathan M. Roberts, and Mandyam V. Srinivasan. 2014. On the use of optical flow for scene change detection and description. Journal of Intelligent & Robotic Systems 74 (3–4): 817–846.CrossRefGoogle Scholar
  7. 7.
    Kim, Giseok, and Jae-Soo Cho. 2012. Vision-based vehicle detection and inter-vehicle distance estimation for driver alarm system. Optical Review 19 (6): 388–393.CrossRefGoogle Scholar
  8. 8.
    Jen, Cheng-Lung, Yen-Lin Chen, and Hao-Yuan Hsiao. 2017. Robust detection and tracking of vehicle taillight signals using frequency domain feature based adaboost learning. In IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW), 423–424.Google Scholar
  9. 9.
    Li, Yi, Zi-xing Cai, and Jin Tang. 2012. Recognition algorithm for turn light of front vehicle. Journal of Central South University 19: 522–526.CrossRefGoogle Scholar
  10. 10.
    Tong, Jian-jun and Zou Fu-ming. 2005. Speed measurement of vehicle by video image. Journal of Image and Graphics 10(2): 192–196.Google Scholar
  11. 11.
    Casares, Mauricio, Akhan Almagambetov, and Senem Velipasalar. 2012. A robust algorithm for the detection of vehicle turn signals and brake lights. In IEEE ninth international conference on advanced video and signal-based surveillance (AVSS), 386–391.Google Scholar
  12. 12.
    Song, Hua-jun, and Mei-li Shen. 2011. Target tracking algorithm based on optical flow method using corner detection. Multimedia Tools and Applications 52 (1): 121–131.CrossRefGoogle Scholar
  13. 13.
    Enkelmann, Wilfried. 1988. Investigations of multigrid algorithms for the estimation of optical flow fields in image sequences. Computer Vision, Graphics, and Image Processing 43 (2): 150–177.CrossRefGoogle Scholar
  14. 14.
    Tom, Mitchell. 1997. Machine learning. McGraw-Hill computer science series.Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  • J. Arunnehru
    • 1
    Email author
  • H. Anwar Basha
    • 1
  • Ajay Kumar
    • 1
  • R. Sathya
    • 2
  • M. Kalaiselvi Geetha
    • 2
  1. 1.Department of CSESRM University (Vadapalani Campus)ChennaiIndia
  2. 2.Department of CSEAnnamalai University, AnnamalainagarChennaiIndia

Personalised recommendations