Computer Vision Toolbox
lastmod 2025-04-21created 2025-04-211 edits
Design and test computer vision systems
About
Computer Vision Toolbox provides algorithms and apps for designing and testing computer vision systems. You can perform feature detection, feature extraction, and feature matching, as well as visual inspection, object detection, and object tracking. You can automate calibration workflows for single, stereo, and fisheye cameras. For 3D vision, the toolbox supports stereo vision, point cloud processing, structure from motion, and real-time visual SLAM and point cloud SLAM. With the computer vision apps you can perform camera calibration as well as team-based ground truth labeling using automation.
You can use pretrained object detectors, or train custom detectors using deep learning and machine learning algorithms such as YOLO, SSD, and ACF. For semantic segmentation and instance segmentation, you can use deep learning algorithms such as U-Net, SOLO, and Mask R-CNN. You can perform image classification using vision transformers such as ViT. With pretrained models you can detect faces and pedestrians, perform OCR (optical character recognition), and recognize other common objects.
You can accelerate algorithms by running them on multicore processors and GPUs. Toolbox algorithms support C/C++ code generation for integration with existing code, desktop prototyping, and embedded vision system deployment.