• What if your cameras could scan the room and immediately tell you how many people are in the standing-room-only crowd?

• Imagine if the city transportation department had cameras counting in real-time the number of people waiting for the #22 bus, and could immediately dispatch additional buses to accommodate rush hour and unexpected surges?

• Consider how much more quickly and easily ranchers could count livestock in a given area to better track and manage their assets?

Building a way to count - fast, accurate and effective

Mobiquity designed and engineered a crowd counting mobile application for AWS, to prove out the combined power of mobile, video streaming to Kinesis Video Streams, and the machine learning capabilities of SageMaker. The result is a working mobile application, which allows a user to scan a room and count the number of people, determined by an object detection model which is trained to recognize 'heads and shoulders' patterns.

This can be modified to recognize different shaped objects - such as cars or livestock.  There are multiple use cases for which this combination of technology can be effective.