The client requested the solution that will help to monitor outdoor ad impressions.
We developed a solution, which is designed to monitor the car traffic moving near the installed IP traffic camera. It tracks ridership (number of ad views or impressions) passing by particular billboards (outdoor ad) per day.
The solution consists of three parts - a common user interface (web page), a subsystem for video stream recognition (client), a subsystem for collecting and analyzing data (server).
The main content of the user page - addresses of traffic cameras with hyperlinks. The latter provided users with video content with detected cars and amount of passengers. Also, users are able to track and monitor data updates each defined period of time (per day, or per month).
The server receives metadata, alerts about it, and displays the user interface. On a certain page of a particular video stream, the video stream and the number of cars received from the User are displayed on the web page.
The designed technology helps the Company to identify the most profitable locations with billboards and make decisions on pricing.
Technology stack:
Python3, Flask, Tensorflow, YOLOv4, Cuda, Imagezmq, OpenCV, Numpy, Scipy.