Knowing that the number of people passing through the doors is one very basic metric. Retailers would hire people to stand at the entry ways to stores with counters and click people off as they entered. This primitive data collection was expensive, unreliable to maintain and usually not very accurate. Automatic people counting answered a need to improve basic business intelligence gathering.
Real-Time door counting systems started to appear over the last 8 or 9 years. One thesis by MA Engineering candidate, Damien Lefloch at the Department of Computer Science, University of Bourgogne (France), entitled “Real-Time People Counting System Using Video Camera” was published in 2007.
Early real time people counting systems used sensors like laser beam detectors, infra-red and thermal sensors with different levels of accuracy and different advantages and setbacks.
Lefloch makes the point that video camera people counting is the most accurate method for counting people. Lefloch developed a system for reducing the overhead images of people passing the wide-angle lens of the camera into blocks, then counting them by digitally sensing the boundaries of the block images as they pass through the field of the camera.
An article dated 2012 appearing in the International Journal of Modern Engineering Research entitled “Segmentation and Counting of People” (Akhil Khare, Kanchan Warke, and Akhilesh Upadhayay) describes a door counting system that used Open CV (Open Source Computer Vision Library) software developed by Intel to create a sophisticated people counting network.
The PCount app comes from several years of software engineering development. It is compatible with any conventional Axis surveillance cameras, directly running on the camera hardware. The software provides analysis and metrics from the camera images and integrates the metrics across all the cameras in the facility, storing the data in the cloud. The system can provide a complete map of customer traffic patterns.
The camera based footfall counting system is relatively new and advanced technology that can help with merchandising and evaluating market potentials within retail stores or traffic patterns in commercial organizations. PCount can help. Please contact us to discuss your needs.