Traffic sign recognition application using scale invariant feature transform method and support vector machine based on android
Ranti Ratnasari, Budhi Irawan, Casi Setianingsih · 2017
Although there are a lot of traffic signs that exist, but oftentimes people disregard it for various reasons, such as not seeing the existence of these signs and do not know the meaning of signs that exist. It has an impact on traffic order. With this problem, in this final assignment will be designed a mobile application based on android that can introduce traffic signs in different ways. Applications designed using the Scale Invariant Feature Transform (SIFT) and Support Vector Machine (SVM) methods that can provide users with information on the name of the signs and a brief explanation of the sign in Bahasa and English. Based on the results, this application has the highest accuracy rate of 88%. This shows that application can detect 22 of 25 signs. The value obtained when the image taken on 10896 lux - 4105 lux with the position of the image on the Range of Interest (ROI) that contained in the camera.