Visitor Counting and Identification System using Facial Recognition

Project Title

Visitor Counting and Identification System using Facial Recognition

Project Area of Specialization Artificial IntelligenceProject Summary

Visitor counting and Identification system (VCIS) is a web-based system that will use a Facial-recognition technique to count the number of visitors in a shopping mall. It will keep a record of the number of visitors including time-spent by each visitor. It will classify the visitors into 3 categories, i.e. ‘Former visitor’, ‘New visitor’ and ‘Frequent visitor’. Furthermore, a distinction between employees and visitors will be made. The data will be used to predict the number of visitors for the future. The system will assist the shop owners and mall administration to plan ahead for future congestion hours.

The main features are:

• Live streaming of the visitors’ arrival and departure
• Count and classify visitors
• Distinguish between visitors and employees
• Forecast the number of visitors
• Calculate the time spent by each visitor
• Determine the congestion hours

Project Objectives

• Purpose of the proposed VCIS is to facilitate the business owner about the insight that whether the business is growing or lagging, keeping in view that our system will determine the number of visitors visiting the shopping mall.
• Another primary objective is to classify among old, frequent and new visitors which will aid in determining whether the business is growing (i.e. if there are new visitors) or lagging (if the number of visitors are reducing).
• Our product will forecast the number of visitors in the upcoming time which will facilitate the owner to make necessary arrangements and it will be a source of business intelligence for the owner.
• Our product will estimate the congestion hours using the data-analytics techniques on the previously maintained dataset.

Project Implementation Method

We will be using Django framework along with Python for the implementation of this Project. We'll place two cameras in the shopping mall, one at the entrance and one at the exit. The camera at the entrance will first detect a face and then perform matching through a facial recognition technique and consequently assign a unique id to it and store the cropped face, if the face isn't already present in the DB. If the face is already present in the DB, our system will classify among 'Former visitor' and 'Frequent visitor' using the details about when that particular visitor last visited which would be stored in the Database. This whole computational process will take place on a server. Whereas, the client end will only be responsible for sending the live stream (Frames). Using techniques such as the Time-Series approach, we will predict the number of visitors in the upcoming time. Time-spent by each visitor will also be computed using the time difference between arrival and departure time. Also, we will determine congestion hours using the time-stamp stored in the database. Furthermore, a distinction between employees and visitors will also be made using a classification technique

Benefits of the Project

Benefits of our system will be as follows:

• We will be able to know the number of victims if a building collapses in case of a disaster such as an Earthquake
• Our system will provide business insight to the business owner such as number of visitors to a shopping mall
• Classify visitors into three categories (Former, new and frequent visitors)
• Will predict the number of visitors in the future using Forecasting techniques
• Determining congestion hours so work-force may be managed accordingly
• Our system will identify suspicious visitor based on the number of visits in a specific time

Technical Details of Final Deliverable

Our system will use the following technologies:

• Python3+
• Django Framework
• PostgreSQL
• OpenCV
• TensorFlow/Keras
• Numpy/Pandas/Matplotlib
• FaceNet Model
• Caffenet model (DNN)

Final Deliverable of the Project HW/SW integrated systemCore Industry OthersOther Industries IT , Security Core Technology Artificial Intelligence(AI)Other Technologies OthersSustainable Development Goals Decent Work and Economic Growth, Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	70000
Cameras	Equipment	2	8000	16000
DVR	Equipment	1	5000	5000
Banner and standee	Miscellaneous	1	6000	6000
Printing	Miscellaneous	3	1000	3000
Server	Equipment	1	40000	40000