Kenneth Dawson-Howe's Computer Vision course.
The problem is to develop a program in C++ to locate and recognise 81 information signs from Dublin train stations in the 33 provided images.
- This project uses openCV v3.2. Unix download is available from OpenCV.
- CMake available from CMake.
- In order to
#include filesystem
with Clang compiler you wil need Boost library available from Boost. - You may have to add
export OpenCV_DIR=/usr/local/share/OpenCV
to yourbash_profile
so that CMake will know where to find OpenCV libraries.
-
Make a build directory and navigate to it to keep your workspace clean.
mkdir build
cd build
-
cmake ..
-
Build the binary executable.
make
-
To run the program, you must be in the base directory (i.e. the one containing the "Blue Signs" folder)
cd ..
build/blueSignRecognition
ObjectAndLocation::compareObjects()
perfroms recognition of blue signs.ImageWithBlueSigns::LocateAndAddAllObjects()
performs detection of blue signs.