Crowd Counting

Dataset of crowd photos with keypoints labeling of each person

Case Description

Data was obtained by parsing photos of protests, concerts, and other mass events from the internet

: – There is no violence in the photos
– 99% of the photos were taken in daylight

: – Each person is labeled with a keypoint
– Areas where people cannot be clearly observed are annotated

: – Labeling is a set of coordinates in a JSON-file
– Images are classified by the number of people in the crowd

Original image

Labeling of the image

01.
Crowd control at mass events:
Object Detection and Computer Vision for determining crowd density and movement, and counting the number of people for safety and event planning
02.
Civil planning and emergency management:
Computer Vision for determining the most effective evacuation routes
03.
Crowd behavior analysis:
Computer Vision and Classification for analyzing crowd behavior in different situations
04.
Visitor counting:
Object Detection for counting visitors in public places, determining peak visitation times, analyzing popularity, and planning necessary resources and services for visitors