COVIDSafe

Project Title

COVIDSafe

Project Area of Specialization Artificial IntelligenceProject Summary

Due to the increase in covid-19 cases, many people got infected just because the government cannot do lockdowns for much longer and open all the businesses and institutes. Even after all these cases and deaths, not everyone follows the SOP's, and due to that, others got infected.

The project will be used to detect and monitor if people are following the SOP's regarding the Covid-19 situation. The project will help out the mall, institutes, and other commercial areas to detect if people are doing social distancing properly or wearing the mask. If the people do not follow the SOP's, the system will get alert and show violations so the related security department can check out the people and let them follow the SOP's properly. The main idea for the project is that because many people are not following SOP's properly, it is dangerous to others' lives also, so to make sure that people are safe while being in malls or any rush areas, we are designing a system that will help to reduce the spread of coronavirus.

Project Objectives

To design a real-time detection system that detects the pedestrian's, social distance, face, mask, and show a violation counter using a camera, which helps to reduce the spread of Covid-19.
 

Project Implementation Method

The System will use python as the main language; for the algorithm, we will be using Yolov3 (You only Look once), which is a deep learning object detection model, YOLO is an ingenious convolutional neural network (CNN) for doing object detection in real-time. The algorithm applies a single neural network to the full image, and then divides the image into regions and predicts bounding boxes and probabilities for each region. It is super-fast than most of the other algorithms out there, and it is nearly as accurate as SSD (Single-Shot detection). SSD divides the image using a grid and have each grid cell be responsible for detecting objects in that region of the image. Detection objects mean predicting the class and location of an object within that region. For libraries, we are using NumPy for adding support for large, multidimensional arrays and matrices, along with an extensive collection of high-level mathematical functions to operate on these arrays. OpenCV For real-time object Detection, time, math, sys, and imutils.

For GPU access, the System will be using Cuda technology for a parallel computing platform and application programming interface model created by Nvidia. To run Cuda, we need Cuda 11.1 development Kit and cuDNN 8.0, which is NVIDIA CUDA® Deep Neural Network library (cuDNN) is a GPU-accelerated library of primitives for deep neural networks. cuDNN provides highly tuned implementations for standard routines such as forward and backward convolution, pooling, normalization, and activation layers, which is the requirement for Yolov3.

Benefits of the Project

As we all know, According to WHO (World health organization), Covid-19 is not getting a perfect cure anytime soon; WHO has address that we still need at least two years for this Covid-19 situation to be under control, but it will still be here just like SARS disease. We saw that due to people who are not following SOP’s are dangerous to others; even in our country, the cases dramatically increased. So our system will help us detect people who are not following the SOP’s so the related security department of the mall, institutes, etc., can check on them and tell them to follow the SOP’s, which will significantly reduce the spread of Covid19 cases.

Technical Details of Final Deliverable

The final project system will contain software and GPU hardware to run that software. The software will have two options, to run it on live CCTV footage or run it on a pre-recorded footage. Once the user selects any one of that option and upload the CCTV live credentials or upload a pre-recorded video, the software will start the detection and will show the red boxes for people who are not wearing the mask or not doing social distance, simultaneously it will show green boxes for people who are wearing the mask and doing social distance. The user can also see the total number of people in the frame and the total number of violations done by people. All this information will be saved in excel so a data analyst can apply data science to that data and can provide some valuable information in the future.

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 PeopleRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	70000
Nvidia RTX 2060 6GB Edition	Equipment	1	70000	70000