visp

Why we included this project

Visual servoing is a niche that most computer vision libraries only touch as a side feature, but ViSP is built for it from the ground up, closing a control loop on live image measurements to steer a robot. If your work involves robot arms, mobile platforms, or camera-guided hardware, you get model-based object tracking and 3D pose estimation plus the control laws that turn those visual features into actuator commands. The same tracking and pose routines also drive real-time augmented reality and computer animation, so a purely robotic deployment is not the only route. It is a modular C++ codebase with Python bindings and CMake builds, with ROS2 coverage in the build matrix, which makes slotting in your own algorithms more practical than writing a servoing stack from zero. The built-in simulation lets you test a vision-driven controller before your hardware is ready.