Skip to content
Simple exercise solution. Convolutional neural net classifier applied to the image by the sliding window method. Using c++, OpenCV, dlib.
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Classifiers.cpp
Classifiers.h
CommonDefinitions.h
Descriptors.cpp
Descriptors.h
README.md
RandomWarpedImage.h
Resnet.h
TDNNClassifier.h
TData.cpp
TData.h
TDataset.cpp
TDataset.h
TIndexedObjects.h
TRandomDouble.h
TRandomInt.h
TTrafficSignClassifier.cpp
TTrafficSignClassifier.h
TTrafficSignDataset.cpp
TTrafficSignDataset.h
WriteToLog.h
cascade_detector.cpp
cascade_detector.h
main.cpp
multiscale_sampler.cpp
multiscale_sampler.h

README.md

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.

You can’t perform that action at this time.