Real Time Crowd Counting COVID SOPs Detection System

Project Title

Real Time Crowd Counting COVID SOPs Detection System

Project Area of Specialization Artificial IntelligenceProject Summary

Social distancing has been proven as an effective measure against the spread of the infectious Coronavirus Disease. However, individuals are not used to tracking the required 3 feet distance between themselves and their surroundings, also businesses have to comply with restrictions on the number of visitors based on the new government social distancing regulations. Having to control the number of visitors poses a problem for essential public service providers. To ensure that people are as spaced out as needed requires an active surveillance system capable of detecting distances and total number of individuals in a particular place. Also to detect whether people are wearing face mask or not will be our next objective. So we propose an Artificial Intelligence (AI) based real-time crowd counting, SOPs detection and warning system.

Project Objectives

Objectives are:

-- Detect humans separately from other creatures.

-- Count humans in a particular area.

-- Must detect mask on faces

-- Calculate distance between two humans

-- Detect whether distance is less than 3 feet, or if someone is not wearing mask.

--  System must notify concerned authorities, who will be able to get useful insights, and take actions according to it.

Project Implementation Method

For this Artificial Intelligence (AI) based real-time crowd counting, SOPs detection and warning system we consider two important ethical factors:

1. The system should never record/cache data.

2. The warnings should not target the individuals.

-- The video stream input will be taken continuously through the deployed CCTVs at public places considering scope of our project.

-- The input video will be sent to our AI based system that will perform its tasks, such as:

-- Extract frames from the video input

--- Detection of ‘Human Class’ objects only (Image Processing).

-- Applying deep convolutional neural networks, different Deep Learning and Machine Learning algorithms in order to produce results.

-- Find out number of people observed using ‘head counting technique’ in those frames and sum them up.

-- See for violations of SOPs. 

-- Check people captured in those frames, are their faces covered or not? Count the number of people whose faces are not covered. This will require such an algorithm which will detect the presence of cloth above nose.

-- Then check for the number of violations in those frames related to social distancing.

If people do not comply with 3 feet distance even after 1 minute, then this would be considered as violation.

-- After getting these results the crowd count will be notified to the concerned authority.

-- And depending on the above conditions if SOPs are violated then conecerned authority will be notified. Also admin would be able to check logs and details for particular times. This will make people remind and aware to follow SOPs.

Benefits of the Project

We aimed to work on the current global problem of pandemic which is highly concerned at this time which is basically maintaining the SOPs among the individuals. The idea of our internal advisor and our personal motivation to solve this problem compelled us to work on this project. The application which is to be developed will help the government and the concerned authorities to take immediate action against the individuals not following the SOPS as our system will immediately notify them in real time.

Technical Details of Final Deliverable

Desktop Application that can take input from the camera in real time, to be used by the admin. 

Final Deliverable of the Project HW/SW integrated systemCore Industry ITOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Good Health and Well-Being for People, Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	70000
Original Raspberry PI 4 Model B with 4GB RAM	Equipment	1	20000	20000
GPU Card	Equipment	1	15000	15000
Drone DJI spark	Equipment	1	35000	35000