2016 International Conference on Frontiers of Information Technology (FIT) (2016)
Dec. 19, 2016 to Dec. 21, 2016
Didit Andri Jatmiko , School of Electrical Engineering and Informatics, Bandung Institute of Technology, Bandung, Indonesia
Ary S. Prihatmanto , School of Electrical Engineering and Informatics, Bandung Institute of Technology, Bandung, Indonesia
In this paper, we present a system to recognize text in traffic signs, along with its context based recognition result corrections that we developed. This system detects text in traffic signs region using contour detection and using KNN Classifier to recognize letters in it. The result of the recognitions that may contain errors will be corrected using Forward Reverse Dictionary that has Contextual Database. This testing is done for recognition system without correction and recognition system with correction on a sample sign. This implementation increases the accuracy rate of word recognition by 10% at 10% noise on a 100% area which is quite good.
Context, Traffic Signs Text, KNN, Forward Reversed Dictionary, Contextual Database, Word Relations,
Didit Andri Jatmiko, Ary S. Prihatmanto, "Traffic signs text recognition and error correction", 2016 International Conference on Frontiers of Information Technology (FIT), vol. 00, no. , pp. 123-127, 2016, doi:10.1109/FIT.2016.7857550