icaroalvarez / cv-pipeline

Visually testing image processors

CV-Pipeline - Computer Vision Pipeline.

Process images through a pipeline of image processors. Configure and tune the processors and see the result in real time.

Installing external dependencies

Dependencies managed by conan

• Catch2
• Trompeloeil
• OpenCV
• JSON for Modern C++
• Easylogging++

Run the following script for installing the dependencies:

mkdir .conan && cd .conan
conan install ..

Building the project

Using CMake

To compile any target in the project, you can directly use CMake. Create a build directory manually and generate the Makefiles in it.

mkdir build && cd build
cmake ..
make

Using image processor pipeline

• Create your own custom image processors or use the ones already created.
• Execute pipeline processing

Technical Features

• Modern C++17 (std::variant, std::visit).
• Testing and mocking with Catch2 and Trompeloeil frameworks.
• Testing coverage using lcov.
• Acceptance Test Driven Development (ATDD) and Behavior-Driven Development (BDD).
• Build system with Modern CMake.
• Conan as dependency manager.
• Multi-threading synchronization with condition variables.
• Template specialization