How It Works - The Technology
A camera is placed inside the store, in front of the front doors in a store. As and when shoppers come into the store they are detected by the camera and then counted be in into the store and then into a database. A high res camera with WDR and intelligent sensing mechanisms is placed inside the store facing the front doors at eye level or just thereabouts
One standard IP zoom camera will be stationed at the front door. Its sole purpose is to take continuous video of everyone walking in. IConnect then takes the video file, strips out the images and runs an algorithm to determine if each subject is male or female, adult or child, walking in alone or in a group of two, three or more.
After that, it runs several more algorithms for facial recognition, and through this, it's able to tell if the face in the new video matches the faces of any others who have walked through the doors over the past 90 days. The camera will try to take a face picture and if not, then do a face detection. The system has in its memory 80 different measures of algorithms that will compare face pictures to suggest a person walked into the store. Facial recognition & comparisons is done with pictures stored in the database against previous visits and then it creates the metrics for reporting and analytics.
Use of this technology will overcome the issue of accurate shopper counts in a store by the hour in a given day and then be able to compute sales per guest & close ratio percentage.
Use of this technology will enable retailers to know how many people come in every day, every hour, week day v/s weekends or holidays, repeat v/s first time visitors, buying groups, gender determination, age determination, moods of the customers when in the store and when planning on leaving the store. The camera technology will provide the ability to generate heat maps for retailers to identify what part of the store sells and what part is being frequented most and where do people spend the majority of the time.