Neural Computing and Applications

, Volume 28, Supplement 1, pp 573–584 | Cite as

Embedded real-time speed limit sign recognition using image processing and machine learning techniques

  • Samuel L. Gomes
  • Elizângela de S. Rebouças
  • Edson Cavalcanti Neto
  • João P. Papa
  • Victor H. C. de Albuquerque
  • Pedro P. Rebouças Filho
  • João Manuel R. S. Tavares
Original Article


The number of traffic accidents in Brazil has reached alarming levels and is currently one of the leading causes of death in the country. With the number of vehicles on the roads increasing rapidly, these problems will tend to worsen. Consequently, huge investments in resources to increase road safety will be required. The vertical R-19 system for optical character recognition of regulatory traffic signs (maximum speed limits) according to Brazilian Standards developed in this work uses a camera positioned at the front of the vehicle, facing forward. This is so that images of traffic signs can be captured, enabling the use of image processing and analysis techniques for sign detection. This paper proposes the detection and recognition of speed limit signs based on a cascade of boosted classifiers working with haar-like features. The recognition of the sign detected is achieved based on the optimum-path forest classifier (OPF), support vector machines (SVM), multilayer perceptron, k-nearest neighbor (kNN), extreme learning machine, least mean squares, and least squares machine learning techniques. The SVM, OPF and kNN classifiers had average accuracies higher than 99.5 %; the OPF classifier with a linear kernel took an average time of 87 \(\upmu\)s to recognize a sign, while kNN took 11,721 \(\upmu\)s and SVM 12,595 \(\upmu\)s. This sign detection approach found and recognized successfully 11,320 road signs from a set of 12,520 images, leading to an overall accuracy of 90.41 %. Analyzing the system globally recognition accuracy was 89.19 %, as 11,167 road signs from a database with 12,520 signs were correctly recognized. The processing speed of the embedded system varied between 20 and 30 frames per second. Therefore, based on these results, the proposed system can be considered a promising tool with high commercial potential.


Cascade haar-like features Pattern recognition Computer vision Automotive applications 



Pedro Pedrosa Rebouças Filho acknowledges the sponsorship from the Instituto Federal do Ceará (IFCE) via grants PROINFRA/2013, PROAPP/2014 and PROINFRA/2015. Victor Hugo C. Albuquerque acknowledges the sponsorship from the Brazilian National Council for Research and Development (CNPq) through grants 470501/2013-8 and 301928/2014-2. Authors gratefully acknowledge the funding of Project NORTE-01-0145-FEDER-000022, SciTech—Science and Technology for Competitive and Sustainable Industries, co-financed by Programa “Operacional Regional do Norte (NORTE2020)” through “Fundo Europeu de Desenvolvimento Regional (FEDER).”


  1. 1.
    Albuquerque VHC, Barbosa CV, Silva CC, Moura EP, Rebouças Filho PP, Papa JP, Tavares JMRS (2015) Ultrasonic sensor signals and optimum-path forest classifier for the microstructural characterization of thermally-aged inconel 625 alloy. Sensors 15(6):12,474CrossRefGoogle Scholar
  2. 2.
    Albuquerque VHC, Rebouças Filho PP, da Silveira Cavalcanti T, Tavares JMRS (2010) New computational solution to quantify synthetic material porosity from optical microscopic images. J Microsc 240(1):50–59MathSciNetCrossRefGoogle Scholar
  3. 3.
    Amat F, Keller P (2013) 3D Haar-like elliptical features for object classification in microscopy. In: 10th international symposium on biomedical imaging (ISBI), pp 1194–1197Google Scholar
  4. 4.
    Arbib MA (2003) The handbook of brain theory and neural networks. MIT Press, CambridgezbMATHGoogle Scholar
  5. 5.
    de Azevedo FM, Brasil LM, de Oliveira RCL (2000) Neural networks with applications control and expert systems. Visual BooksGoogle Scholar
  6. 6.
    Barreto G, Frota R (2013) A unifying methodology for the evaluation of neural network models on novelty detection tasks. Pattern Anal Appl 16(1):83–97MathSciNetCrossRefGoogle Scholar
  7. 7.
    Barros ALBP, Barreto GA (2012) Extreme learning machine robusta para reconhecimento de faces. In: Brazilian conference on intelligent systems. Curitiba, PR, BrasilGoogle Scholar
  8. 8.
    Barthès JPA, Bonnifait P (2015) Chapter 9 - Multi-Agent active collaboration between drivers and assistance systems. In: Advances in artificial transportation systems and simulation, pp 163–180. Academic Press, BostonGoogle Scholar
  9. 9.
    Bittencourt G (2006) Artificial Intelligence - Tools and Theories, 3 edn. Federal University of Santa CatarinaGoogle Scholar
  10. 10.
    Burges C (1998) A tutorial on support vector machines for pattern recognition. Data Mining Knowl Discov 2(2):121–167CrossRefGoogle Scholar
  11. 11.
    Canny J (1986) A computational approach to edge detection. IEEE Trans Pattern Anal Mach Intell 6:679–698CrossRefGoogle Scholar
  12. 12.
    Carrese S, Mantovani S, Nigro M (2014) A security plan procedure for heavy goods vehicles parking areas: an application to the lazio region (Italy). Transp Res E Logist Transp Rev 65:35–49CrossRefGoogle Scholar
  13. 13.
    Chang CC, Lin CJ (2011) Libsvm: a library for support vector machines. ACM Trans Intell Syst Technol 2(3):27:1–27:27CrossRefGoogle Scholar
  14. 14.
    Cortes C, Vapnik V (1995) Support vector networks. Mach Learn 20(3):273–297zbMATHGoogle Scholar
  15. 15.
    Duda RO, Hart PE (1972) Use of the hough transformation to detect lines and curves in pictures. Commun ACM 15(1):11–15CrossRefzbMATHGoogle Scholar
  16. 16.
    da Silva Felix JH, Cortez PC, Rebouças Filho PP, de Alexandria AR, Costa RCS, Holanda MA (2008) Identification and quantification of pulmonary emphysema through pseudocolors. In: MICAI 2008: Advances in Artificial Intelligence, pp 957–964. SpringerGoogle Scholar
  17. 17.
    Elmer P, Lupp A, Sprenger S, Thaler R, Uhl A (2015) Exploring compression impact on face detection using haar-like features. In: Paulsen RR, Pedersen KS (eds) Image analysis, lecture notes in computer science, vol 9127, pp 53–64. Springer International PublishingGoogle Scholar
  18. 18.
    Falcão AX, Stolfi J, Lotufo RA (2004) The image foresting transform theory, algorithms, and applications. IEEE Trans Pattern Anal Mach Intell 26(1):19–29CrossRefGoogle Scholar
  19. 19.
    Garcia I, Bronte S, Bergasa L, Almazan J, Yebes J (2012) Vision-based drowsiness detector for real driving conditions. In: Intelligent vehicles symposium (IV), pp 618–623Google Scholar
  20. 20.
    Glasbey CA (1993) Analysis of histogram-based thresholding algorithms. CVGIP Graph Models Image Process 55:532–537CrossRefGoogle Scholar
  21. 21.
    Gomes SL, Rebouças ES, Rebouças Filho PP (2014) Reconhecimento Óptico de caracteres para reconhecimento das sinalizações verticais das vias de trânsito. Rev SODEBRAS 9:9–12Google Scholar
  22. 22.
    Haykin SO (2008) Neural networks and learning machines. Pearson Prentice Hall, Upper Saddle RiverGoogle Scholar
  23. 23.
    Helene O (2006) Method of least squares. Livraria da FísicaGoogle Scholar
  24. 24.
    Horata P, Chiewchanwattana S, Sunat K (2013) Robust extreme learning machine. Neurocomputing 102:31–44CrossRefGoogle Scholar
  25. 25.
    Huang GB, Chen L, Siew CK (2006) Universal approximation using incremental constructive feedforward networks with random hidden nodes. IEEE Trans Neural Netw 17:879–892CrossRefGoogle Scholar
  26. 26.
    Huang GB, Wang DH, Lan Y (2011) Extreme learning machines: a survey. Int J Mach Learn Cybern 2:107–122CrossRefGoogle Scholar
  27. 27.
    Huang GB, Zhu QY, Siew CK (2006) Extreme learning machine: theory and applications. Neurocomputing 70:489–501CrossRefGoogle Scholar
  28. 28.
    Kocer HE, Cevik KK (2011) Artificial neural networks based vehicle license plate recognition. Proc Comput Sci 3:1033–1037CrossRefGoogle Scholar
  29. 29.
    Kohonen T (1989) Self-organization and associative memory, 3rd edn. Springer-Verlag New York Inc, New York, NYCrossRefzbMATHGoogle Scholar
  30. 30.
    Lienhart R, Maydt J (2002) An extended set of haar-like features for rapid object detection. In: International conference on image processing, vol 1, pp I–900–I–903Google Scholar
  31. 31.
    McAndrew A (2004) Introduction do digital image processing with matlab. Thomson LearningGoogle Scholar
  32. 32.
    Medeiros C, Barreto G (2013) A novel weight pruning method for mlp classifiers based on the maxcore principle. Neural Comput Appl 22(1):71–84CrossRefGoogle Scholar
  33. 33.
    Mena AP, Bachiller Mayoral M, Díaz-Lópe E (2015) Comparative study of the features used by algorithms based on viola and jones face detection algorithm. In: Bioinspired computation in artificial systems, lecture notes in computer science, vol 9108, pp. 175–183. Springer International PublishingGoogle Scholar
  34. 34.
    Minsky M, Papert S (1969) Perceptrons. MIT Press, CambridgezbMATHGoogle Scholar
  35. 35.
    Moreira FDL, Kleinberg MN, Arruda HF, Freitas FNC, Parente MMV, de Albuquerque VHC, Rebouças Filho PP (2016) A novel vickers hardness measurement technique based on adaptive balloon active contour method. Expert Syst Appl 45:294–306CrossRefGoogle Scholar
  36. 36.
    Neto EC, Gomes SL, Filho PPR, de Albuquerque VHC (2015) Brazilian vehicle identification using a new embedded plate recognition system. Measurement 70:36–46CrossRefGoogle Scholar
  37. 37.
    Neto EC, Rebouças ES, Moraes JL, Gomes SL, Rebouças Filho PP (2015) Development control parking access using techniques digital image processing and applied computational intelligence. IEEE Transactions on Latin. IEEE Trans Latin America 13:272–276CrossRefGoogle Scholar
  38. 38.
    Nissen S (2003) Implementation of a fast artificial neural network library (FANN). Department of Computer Science University of Copenhagen (DIKU)Google Scholar
  39. 39.
    Papa JP, Falcão AX, de Albuquerque VHC, Tavares JMRS (2012) Efficient supervised optimum-path forest classification for large datasets. Pattern Recognit 45(1):512–520CrossRefGoogle Scholar
  40. 40.
    Papa JP, Falcao AX, Suzuki CT (2009) Supervised pattern classification based on optimum-path forest. Int J Imaging Syst Technol 19(2):120–131CrossRefGoogle Scholar
  41. 41.
    Papa JP, Falcão AX, Suzuki CTN (2009) Supervised pattern classification based on optimum-path forest. Int J Imaging Syst Technol 19(2):120–131CrossRefGoogle Scholar
  42. 42.
    Plucker JA, Esping A (2016) Human intelligence: historical influences, current controversies, teaching resources.
  43. 43.
    Rakate G, Borhade S, Jadhav P, Shah M (2012) Advanced pedestrian detection system using combination of haar-like features, adaboost algorithm and edgelet-shapelet. In: IEEE international conference on computational intelligence computing research (ICCIC), pp 1–5Google Scholar
  44. 44.
    Rebouças Filho PP, Cortez PC, da Silva Barros AC, Albuquerque VHC (2014) Novel adaptive balloon active contour method based on internal force for image segmentation - a systematic evaluation on synthetic and real images. Expert Syst Appl 41(17):7707–7721CrossRefGoogle Scholar
  45. 45.
    Rebouças Filho PP, Moreira FDL, de Lima Xavier FG, Gomes SL, Santos JC, Freitas FNC, Freitas RG (2015) New analysis method application in metallographic images through the construction of mosaics via speeded up robust features and scale invariant feature transform. Materials 8(7):3864CrossRefGoogle Scholar
  46. 46.
    Rebouças Filho PP, Cortez PC, Félix JHDS, Cavalcante TdS, Holanda MA (2013) Adaptive 2d crisp active contour model applied to lung segmentation in ct images of the thorax of healthy volunteers and patients with pulmonary emphysema. Revista Brasileira de Engenharia Biomédica 29(4):363–376CrossRefGoogle Scholar
  47. 47.
    Rezaei M, Ziaei Nafchi H, Morales S (2014) Global haar-like features: a new extension of classic haar features for efficient face detection in noisy images. Image and Video Technology, Lecture Notes in Computer Science, vol 8333, pp 302–313. Springer Berlin HeidelbergGoogle Scholar
  48. 48.
    Riedmiller M, Braun H (1993) A direct adaptive method for faster backpropagation learning: the RPROP algorithm. IEEE Int Conf Neural Netw 1:586–591CrossRefGoogle Scholar
  49. 49.
    Ruck DW, Rogers SK, Kabrisky M, Oxley ME, Suter BW (1990) The multilayer perceptron as an approximation to a bayes optimal discriminant function. IEEE Trans Neural Netw 1(4):296–298CrossRefGoogle Scholar
  50. 50.
    Russell SJ, Norvig P (2009) Artificial intelligence: a modern approach, 3rd edn. Prentice Hall, Upper Saddle RiverzbMATHGoogle Scholar
  51. 51.
    Schimidt W (1993) Initialization, backpropagation and generalization of feed-forward classifiers. IEEE Int Conf Neural Netw 1:598–604CrossRefGoogle Scholar
  52. 52.
    Schölkopf B, Smola AJ (2002) Learning with kernels. MIT press, CambridgezbMATHGoogle Scholar
  53. 53.
    Tavares JMR, Rebouças Filho PP, Cavalcante TDS, de Albuquerque VHC (2009) Brinell and vickers hardness measurement using image processing and analysis techniques. J Test Eval 38(1):1–7Google Scholar
  54. 54.
    Tu C, van Wyk B, Hamam Y, Djouani K, Du S (2013) Vehicle position monitoring using hough transform. Int Conf Electron Eng Comput Sci (EECS 2013) 4:316–322Google Scholar
  55. 55.
    Vapnik VN (1999) An overview of statistical learning theory. IEEE Trans Neural Netw 10(5):988–999CrossRefGoogle Scholar
  56. 56.
    Viola P, Jones M (2001) Rapid object detection using a boosted cascade of simple features. IEEE Comput Soc Conf Comput Vision Pattern Recognit 1:511–518Google Scholar
  57. 57.
    Widrow B (1990) 30 years of adaptive neural networks: perceptron, madaline, and backpropagation. Proc IEEE 78:1415–1442CrossRefGoogle Scholar
  58. 58.
    Widrow B, Winter R (1988) Neural nets for adaptative filtering and adaptative pattern recognition. IEEE Comput 21:25–39CrossRefGoogle Scholar
  59. 59.
    Wu BF, Huang HY, Chen CJ, Chen YH, Chang CW, Chen YL (2013) A vision-based blind spot warning system for daytime and nighttime driver assistance. Comput Electr Eng 39(3):846–862CrossRefGoogle Scholar
  60. 60.
    Yi SC, Chen YC, Chang CH (2015) A lane detection approach based on intelligent vision. Comput Electr Eng 42:23–29CrossRefGoogle Scholar
  61. 61.
    Yu S, Shi Z (2015) The effects of vehicular gap changes with memory on traffic flow in cooperative adaptive cruise control strategy. Phys A Stat Mech Appl 428:206–223CrossRefGoogle Scholar
  62. 62.
    Yuen HK, Illingworth J, Kittler J (1989) Detecting partially occluded ellipses using the hough transform. Image Vis Comput 7(1):31–37CrossRefGoogle Scholar
  63. 63.
    Zhang S, Bauckhage C, Cremers A (2014) Informed haar-like features improve pedestrian detection. In: IEEE conference on computer vision and pattern recognition (CVPR), pp 947–954Google Scholar
  64. 64.
    Zheng K, Zhao Y, Gu J, Hu Q (2012) License plate detection using haar-like features and histogram of oriented gradients. In: IEEE international symposium on industrial electronics (ISIE), pp 1502–1505Google Scholar

Copyright information

© The Natural Computing Applications Forum 2016

Authors and Affiliations

  • Samuel L. Gomes
    • 1
  • Elizângela de S. Rebouças
    • 1
  • Edson Cavalcanti Neto
    • 1
  • João P. Papa
    • 2
  • Victor H. C. de Albuquerque
    • 3
  • Pedro P. Rebouças Filho
    • 1
  • João Manuel R. S. Tavares
    • 4
  1. 1.Laboratório de Processamento Digital de Imagens e Simulação Computacional, Instituto Federal de Federal de EducaçãoCiência e Tecnologia do Ceará (IFCE)CearáBrazil
  2. 2.Departamento de Ciência da ComputaçãoUniversidade Estadual PaulistaBauru, São PauloBrazil
  3. 3.Programa de Pós-Graduação em Informática Aplicada, Laboratório de BioinformáticaUniversidade de FortalezaFortalezaBrazil
  4. 4.Instituto de Ciência e Inovação em Engenharia Mecânica e Engenharia Industrial, Departamento de Engenharia Mecânica, Faculdade de EngenhariaUniversidade do PortoPortoPortugal

Personalised recommendations