A ROS-based object detection and pose estimation library for 2D and 3D applications using OpenCV.
Find-Object is an open-source computer vision library for detecting objects and estimating their 3D pose in robotic applications. It uses OpenCV for feature matching and integrates with ROS to process image streams from cameras, enabling robots to recognize and locate objects in their environment.
Robotics developers and researchers working on perception tasks such as object recognition, visual servoing, or augmented reality in ROS-based systems.
Developers choose Find-Object for its straightforward integration with ROS, support for both 2D detection and 3D pose estimation, and reliance on proven OpenCV algorithms without the complexity of deep learning setups.
Find-Object project
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Works with both ROS1 and ROS2, with clear installation and launch commands for each, as shown in the README's binary and source build sections.
Supports RGB-D cameras like Intel RealSense to compute object position and orientation, essential for robotic manipulation and augmented reality tasks.
Leverages OpenCV's feature detection modules, including non-free algorithms, for reliable 2D object recognition without deep learning dependencies.
Builds on Linux and Windows with CI/CD badges indicating tested compatibility, ensuring broader deployment options.
Requires full ROS setup, specific camera drivers, and environment configuration, making initial integration steep compared to lightweight libraries.
Depends on OpenCV's xfeatures2d and nonfree modules, which may have licensing issues and require manual installation, adding overhead.
Designed for known objects via feature matching, not for general or large-scale object detection, limiting use in dynamic environments.
The README frequently redirects to an external ROS wiki for parameters and details, which can be fragmented and less accessible for quick reference.