Traffic sign recognition based on the convolutional neural net classifier. Background is assigned to one of the classes to avoid creating a detector. Classifier is applied by sliding window method. The program is written in c++ using OpenCV, dlib libraries. For classifier training images from The German Traffic Sign Detection Benchmark (http://benchmark.ini.rub.de/?section=gtsdb&subsection=dataset) are used.
deliriumv01d / traffic-sign-recognition Goto Github PK
View Code? Open in Web Editor NEWSimple exercise solution. Convolutional neural net classifier applied to the image by the sliding window method. Using c++, OpenCV, dlib.