Blob tracking
lastmod 2026-11-10created 2022-07-122 edits
Blob tracking is an image processing technique that finds connected regions (blobs) in an image whose brightness or color stands out from their surroundings, then matches the same blob across frames to follow its motion. Blobs are separated by background subtraction or binarization, their centers and sizes are obtained with OpenCV's SimpleBlobDetector or contour detection, and nearby blobs are linked to build trajectories. It is used for people counting, touch tables, and simple object tracking.