Computer Vision Concepts
The tasks computer vision solves.
This is an overview — pick a sub-topic below.
Sub-topics (18)
Digital Image Processing
Operate on pixels: filtering, thresholding, edges, morphology and colour spaces.
Image Classification
Assign one label to a whole image ("good board" vs "defective").
Object Detection
Find objects and draw bounding boxes around them, with class labels and confidence.
Image Segmentation: Semantic, Instance and Panoptic
Label every pixel: by class (semantic), by individual object (instance), or both (panoptic).
Object Tracking
Follow detected objects across video frames, keeping consistent IDs.
Face Detection, Recognition and Landmarks
Find faces, identify people, and locate facial key points.
Pose Estimation, Hand Tracking and Gesture Recognition
Locate body/hand key points to understand posture and gestures.
Human Activity Recognition and Video Understanding
Recognise actions and events in video (walking, falling, assembling).
Optical Flow, Depth and Stereo Vision
Estimate motion between frames and distance from one or two cameras.
3D Reconstruction, Visual Odometry and SLAM
Build 3-D maps and estimate camera motion from images — the eyes of robots and drones.
Image Registration and Stitching
Align images (medical scans, satellite tiles) and stitch panoramas.
OCR and Document Understanding
Read text from images and extract structure from documents.
Barcode and QR Code Detection
Decode 1-D barcodes and QR codes from images or camera feeds.
Medical and Satellite Image Analysis
Specialised vision for scans (MONAI, SimpleITK) and Earth observation (TorchGeo, rasterio).
Super-Resolution, Denoising and Inpainting
Enhance image detail, remove noise, and fill in missing regions.
Motion Detection and Background Removal
Detect moving objects by subtracting a background model; cut out subjects from images.
Visual Inspection and Industrial Quality Control
Automatically find defects on products — missing components, solder bridges, scratches.
Autonomous Driving Perception
Detect lanes, vehicles, pedestrians and signs, and fuse camera, LiDAR and radar.
Spotted a mistake or want to add detail? Suggest an improvement — Content Contributors' edits are reviewed by editors before they go live.
Discussion (0)
Sign in or join free to ask questions and discuss this topic.