Deteksi dan Pengenalan Rambu Lalu Lintas di Indonesia Menggunakan RGBN dan Gabor

  • Cahya Rahmad Politeknik Negeri Malang
  • Isna Fauzia Rahmah Politeknik Negeri Malang
  • Rosa Andrie Asmara Politeknik Negeri Malang


Traffic signs detection and recognition has an important role in the development of several
expert systems such as ADAS (Advanced Driver-Assistance Systems) and autonomous steering systems.
This study focused on the detection and recognition process tested on Indonesian traffic signs. In
detecting and recognizing traffic signs there are some common problems encountered, such as damaged
traffic signs, fading colors and fluctuating natural lighting conditions. This could potentially lead to
accidents and traffic violations. Therefore, this study is proposed to solve the problem in three main steps.
The first step of segmenting is based on RGBN thresholding (Normalized RGB). Next detects traffic signs
by processing blobs that have been extracted by the previous process. In this step the regionprops are
used to classify the shape of the beacon by labeling each candidate region and analyzed which is the
closest to the size of the blobs. And the last is the process of recognition of signs. In the process of
recognition of signs used a combination of two methods of feature extraction of Gabor, while for the
classification process using SVM and KNN are tested jointly on the results of feature extraction. This
method is tested on dataset detection and introduction of traffic signs Indonesia obtained by processing
data from the Department of Transportation Malang. Experimental research results of this approach
show the precision and recall about 90,13% and 91,07% respectively in detecting traffic signs, and
93,9% for recognizing process.


How to Cite
RAHMAD, Cahya; RAHMAH, Isna Fauzia; ASMARA, Rosa Andrie. Deteksi dan Pengenalan Rambu Lalu Lintas di Indonesia Menggunakan RGBN dan Gabor. Prosiding Sentrinov (Seminar Nasional Terapan Riset Inovatif), [S.l.], v. 3, n. 1, p. TI13-TI22, nov. 2017. ISSN 2477-2097. Available at: <>. Date accessed: 12 june 2021.