Open AccessRecognition of Pedestrian Traffic Light using Tensorflow And SSD MobileNet V2
Author(s) -
Evan Evan,
Meirista Wulandari,
Eko Syamsudin
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1007/1/012022
Subject(s) - pedestrian , computer science , context (archaeology) , pedestrian detection , artificial intelligence , traffic signal , computer vision , deep learning , transport engineering , real time computing , engineering , geography , archaeology
As a government needs to facilitate the whole public including traffic public facilitate. A pedestrian traffic light should be perceived by pedestrian to safe them to cross the street including someone with the vision deficiency problem. Nowadays, a technology supports human being to build some innovations or application more practical. An innovation which adopt the computer vision and deep learning technology can be built to recognize the pedestrian traffic light which is portable and practical in the context of light computing. The paper aim to evaluate threshold value and data parameters to recognition pedestrian traffic light by Tensorflow and SSD MobileNet V2. The sample are green light and red light on pedestrian traffic light. Based on result, more number of data train than data test and higher of threshold value, the accuracy obtained is more better. Object detection accuracy level reaches 97,98%. On the other result, earphone can produce sounds based on detected objects with output accuracy level reaches 100%.