Covid19 is considered to be the most critical period for all countries. The novel corona virus is spreading in a faster manner. The government has taken various measures in order to avoid the spreading of various by implementing lockdown across the whole country and they spray insecticides in various areas and keep the area clean. Though the government is taking several measures, some people are not cooperating with the rules formulated by the government and they violate the rules and increase the chances for spreading the virus. Many people do not follow the rules such as social distancing from each other, not wearing masks and does not follow lock down rules. This leads to worsen the health condition of the society members. In order to make people to follow the above constraints framed by the government, the proposed system is designed in such a way that the above rules are being followed by people. We have used IBM-Node red and Tello drone service for controlling the drone and visual recognition service for detecting social distance and masks. This is one part of our project. The other part of our project is built using PEGA platform. We build an application for scheduling the inspection process as well as maintaining drones in periodic manner. By using this application, government and police people can get benefitted during covid19 situation.

What it does

The purpose is to inspect whether the people in a public place maintain social distancing or not. It also checks whether every individual is wearing face mask or not. If both are not done, the drone raises alarm and plays warning voice note to the public. In addition, it also carries masks and recognizes people without masks and it instructs people without mask to pick and wear it. The proposed system uses tello drone which is used to perform the inspection process. The IBM visual recognition service with the custom dataset is being embedded in the tello drone. The drone camera detects social distance is maintained or not and whether the people in public are wearing masks or not based on the model trained. The proposed system delivers masks to people who are not wearing masks and tells importance of masks and social distancing. Thus, the proposed system favors the society by saving time and helps in lowering the spread of corona virus.

We gathered custom dataset of 1000 images and partitioned that to training and testing dataset. We annotated images using labelimg tool in which there were four categories such as Mask, Non Mask, Social DIstance, No Social Distance. Then we trained our object detection algorithm with custom dataset and gave sample images for testing the algorithm. It gave predicted results on test image with confidence score. The same annotation has been done in ibm visual recognition service. We created and customized tello drone ui flows in ibm node-red service which consists of tello commands, tello telemetry, tello picture and camera dashboard, mask detection ui. Tello commands gives commands to the drone, tello telemetry displays drone's acceleration, speed levels etc. Tello picture and camera dashboard displays predicted results based on the input image and live video feed provided. Mask Detection ui is used to predict mask and non mask using live video.

Another part of our project is C19-Spector application which is built using pega platform. Our application will help government or police people for scheduling inspection process in a particular area on a particular date and time. After getting approval from higher authority, the respective inspection process will happen. After inspection process is over, the details are recorded by the application and it is stored in pega database for future reference. Three reports are generated such as schedule report, graphical representation of inspection and inspection report. Based on the data recorded, a survey can be made by the government officials and police and they can decide whether to continue inspection in this area or not. This application helps them in taking right decision by scheduling inspection in crowdy areas and recording reports for future purpose.

How we built it

The First part of our application is being built using IBM Services such as IBM-Node-Red, IBM-Cloud, IBM-Watson, IBM-TelloDrone, Deep Learning Algorithms (SSD, Yolo).

The next part of our project is C19 - Spector application which is built using pega platform 8.5.3 version. We created two casetypes one for scheduling inspection and other casetype for maintenance of drones.

Challenges we ran into

We had a very short dataset for our project where we hardly find images through internet. some of the images were relevant and some were irrelevant. Creating custom dataset with more number of images was a great challenge while building our project.

Accomplishments that we're proud of

We have designed two parts of our projects , one part is involves tello drone where it performs inspection process by ibm services and deep learning algorithm.

The other part involves C19 - Spector application where it makes the schedule process in a hierarchical manner by recording details into database and generating the reports which are more important for government officials and police people where they can survey reports and make right decision.

What we learned

By doing one part of our project, we have learnt to use IBM cloud services and deploying our application flow on cloud itself.

By doing other part of our project, we learnt how to build solutions for real world problems by using pega platform 8.5.3. It was a new experience for us to build applications using pega platform for solving real world problems and providing easy solution in a short period of time.

What's next for C19 - Spector

C19-Spector can be used by government officials during this covid19 period for scheduling inspection process. Further, we would like to develop this application for government based on their requirements and help society by building this application.

Built With

+ 3 more
Share this project: