Computer Vision and Robotic Perception

We build computer vision systems that detect, segment and locate objects in camera images and 3D scans. They guide robot arms and mobile robots, and run on their own for tasks such as part inspection. We train models on real or synthetic data and deploy them on the target hardware.

Get In Touch

Computer vision work by BCR: a vision-guided welding cell in Isaac Sim, LiDAR point clouds, ship LiDAR in RViz and a part point cloud aligned with ICP

What we do

Object detection, segmentation and pose estimation with deep learning models
3D vision with point clouds from depth cameras and LiDAR (PCL, Open3D, ICP, RANSAC)
Synthetic data generation in NVIDIA Isaac Sim and Omniverse Replicator
Real-time models sized for the robot's or device's own GPU
Vision-guided picking, welding, docking and inspection

Who it's for

Computer vision services for robotics companies, manufacturers and product teams that need machines to see and measure. Black Coffee Robotics handles computer vision development from data to deployment: object detection, segmentation and pose estimation with deep learning, 3D vision with depth cameras and LiDAR, and synthetic data generation in NVIDIA Isaac Sim when labeled data does not exist.

Tools we use