Database of character numbers of speed limit sign pattern recognition

The developed framework was built using C language for the Android Operating System (OS). The integrated algorithm for the automatic analysis of velocity signs is composed of two main steps: Detection and Recognition, where the first step is to find the desired sign in an image with several objects and the second step is to interpret the information on the sign, i.e. the maximum allowed speed, like folow:

figFluxograma3

The vertical R-19 system for optical character recognition of regulatory traffic signs (maximum allowed speed) according to Brazilian standards developed in this work uses a camera positioned at the front of the vehicle, facing forward:

figApp2

  • Table 1: Amount of existing samples for each character of the numbers from 0 to 9.
Number Number of samples
0 1428
1 1841
2 1879
3 1688
4 1824
5 1569
6 1725
7 1414
8 1650
9 1952
  • Samples are saved as follows: “number_N_S.png”

where:

– N is the character number (N ranges from 0 to 9)
– S represents the sample number of characters N (S ranges from 1 to number of N character samples)

Note 1: The character S 0 can be 1 to 3000, the character 1 can be 1 to 4000, and so on, as shown in Table 1.

Note 2: The file “number_0_19.png” containing the sample 19 of the 0 character of the image database.

  • This framework is being evaluated for publication in Neural Computing and Applications. After published, these images will be released for use and citation of the paper will be available below in order to be properly referenced when using this database.