Question

Write a report on AI based Social distancing monitoring using drone with out human interaction report should contain drone design it specification it's algorithm control system algorithm.minimum 10 A4 size page.No plagiarism is allowed other wise down rate is given

Answer 1

Title Page:

Certificate:

ABSTRACT

A quadcopter can achieve vertical flight in a stable manner and be used to monitor or collect data in a specific region such as Loading a mass. Technological advances have reduced the cost and increase the performance of the low power microcontrollers that allowed the general public to develop their own quadcopter. The goal of this project is to build, modify, and improve an existing quadcopter kit to obtain stable flight, gather and store GPS data, and perform auto commands, such as auto-landing. The project used an Aero quad quadcopter kit that included a frame, motors, electronic speed controllers, Arduino Mega development board, and sensor boards and used with the provided Aero quad software. Batteries, a transmitter, a receiver, a GPS module, and a micro SD card adaptor were interfaced with the kit. The aero quad software was modified to properly interface the components with the quadcopter kit. Individual components were tested and verified to work properly. Calibration and tuning of the PID controller were done to obtain proper stabilization on each axis using custom PID test benches. Currently, the quadcopter can properly stabilize itself, determine its GPS location, and store and log data. Most of the goals in this project have been achieved, resulting in a stable and manoeuvrable quadcopter, with surveillance of people following or not following social distancing norms

1. Introduction
   1. Drone based surveillance

A Drone or Quadcopter is a Vehicles have large potential for performing tasks that are dangerous or very costly for humans. Examples are the inspection of high structures, humanitarian purposes or search-and-rescue missions. One specific type of Drone is becoming increasingly more popular lately: the quadcopter.

When visiting large events or parties, professional quadcopters can be seen that are used to capture video for promotional or surveillance purposes. Recreational use is increasing as well: for less than 50 Euros a small remote-controlled quadcopter can be bought to fly around in your living room or garden. In these situations, the quadcopter is usually in free flight. There is no physical contact between the surroundings and the quad copter and no cooperation between the quadcopters If would have the capabilities to collaborate the number of possibilities grows even further. For example, a group of Drone would be able to efficiently and autonomously search a missing person in a large area by sharing data between. Or, the combined load capacity of a group of quad copters can be used to deliver medicine in remote areas.

This report focuses on the use of a commercially available quadcopter platform, the Drone, to perform a task that requires to inspect the people over an area which are following the social distancing norm which is distance between two individuals should be greater than 1m.

As preliminary step towards the view of collaborating aerial robots the choice was made to perform this task in an indoor scenario where position feedback is present. Starting off with position control, additional controller logic can be implemented to counteract the forces imposed by a mass connected to the quadcopter. The choice is made for the Drone, a generalized approach is chosen where possible to encourage reuse of this research’s outcome and deliverables. A helicopter is a flying vehicle which uses rapidly spinning rotors to push air downwards, thus creating a thrust force keeping the helicopter aloft. Conventional helicopters have two rotors. These can be arranged as two coplanar rotors both providing upwards thrust but spinning in opposite directions (in order to balance the torques exerted upon the body of the helicopter).

Figure 1 Image of Drone Designed by Us

2. Literature review

Various investigations have been directed towards developing an approach that provides a comprehensive solution for crowd management and analysis. These studies have focused on three main principles of crowd analysis including estimation of the density of a group of people per square meter, finding the direction of motion of the crowds and geo-referencing the crowd images in the real-world coordinate system. It can be noticed that most of the studies have been conducted in either computer vision or navigation disciplines. Thus, crowd analysis approaches have been based on images processing and positioning solutions. In this regard, several sensors such as camera (colour, infrared), GPS and IMU (Inertial Management Unit) have usually been mounted on a moving vehicle to detect and analyse the density along with the motion of the crowd in a specific location.[1]

The objective of the research is to classify data finding discriminative features and its analysis. Feature selection weighing based method has been used to discriminate different categories based on the weight of each feature. Among the data of different categories margins are maximized and weight of each feature is determined. The author has proposed Kernel Gram Schmidt process in order to get orthogonal basis set of training and test data in Kernel Space.[2]

The goal of this paper is to implement a system, titled as Drone Map Creator (DMC) using Computer Vision techniques. DMC can process visual information from an HD camera in a drone and automatically create a map by stitching together visual information captured by a drone. The proposed approach employs the Speeded up robust features (SURF) method to detect the key points for each image frame; then the corresponding points between the frames are identified by maximizing the determinant of a Hessian matrix. Finally, two images are stitched together by using the identified points. Our results show that despite some limitations from the external environment, we could have successfully stitched images together along video sequences.[3]

Its derived equations of motion for a quadcopter, starting with the voltage-torque relation for the brushless motors and working through the quadcopter kinematics and dynamics. It ignored aerodynamical effects such as blade-flapping and non-zero free stream velocity but included air friction as a linear drag force in all directions. It used the equations of motion to create a simulator in which to test and visualize quadcopter control mechanisms.[4]

The purpose of this paper is to present the basics of quadcopter modelling and control as to form a basis for further research and development in the area. This is pursued with two aims. The first aim is to study the mathematical model of the quadcopter dynamics. The second aim is to develop proper methods for stabilisation and trajectory control of the quadcopter.[5]

1. Conclusion:

From literature review We can find that the there are two phase for the AI based monitoring 1^st Human detection and 2^nd image analyzation

3. Work Done
   1. MECHANICAL DESIGN:

3 cell 1,000 mAH LiPo rechargeable battery; High pitch propeller for great manoeuvrability; 4 brushless in runner motors with micro ball bearing and rare earth magnets, 14.5 watt & 28,500 rpm when hovering; Self-lubricating bronze bearings, tempered steel prop shafts; Low noise Nylatron gears for 8.625 propeller shafts; Emergency stop controlled by software; Fully reprogrammable motor controller; Water resistant electronic motor controller ; Foam to isolate the inertial centre from the engine’s vibrations; EPP hull; Carbon fibre tubes, 380g with outdoor hull, 420g with indoor hull; High grade 30% fibre charged nylon plastic parts;

1. Material Selection:

Sr.no	Part	Material
1	Frame	CFRP
2	Motors x4	Aluminum Alloy
3	Flight control board	Ceramic
4	Propellers	CFRP
5	Gear	Nylon
6	Battery	Li-ion

2. Computer Programming:

We have programmed the drone on python, Object recognition is a general term to describe a collection of related computer vision tasks that involve identifying objects in digital photographs.

Image classification involves predicting the class of one object in an image. Object localization refers to identifying the location of one or more objects in an image and drawing abounding box around their extent. Object detection combines these two tasks and localizes and classifies one or more objects in an image.

Figure 2 GUI of our algorithm

3. Features of Drone:

HIGH BATEERY LIFE AND FLIGHT TIME

2 KM 1080P VIDEO RECORDING

MAXIMUM SPEED OF 50 KMPH

SMART RETURN TO CHARGING DOCK

LIVE VIDEO FEEDING TO CONTROL ROOM

HIGH CONNECTIVITY AND GPS POSITIONING

EASY TO USE APPLICATION

FOCUS TRACK AND WIDE ANGLE

LOUD BUZZER DUAL SPEAKER

DIRECT NOTIFICATION TO INDIVIDUALS VIA IMAGE PROCESSING

SMART OBSTACLE AVOIDANCE

AUTO CHANNEL SWITCH

Figure 3 Drone

4. Technical Specification of Drone

SENSORS INSIDE SYSTEM

ACCELEROMETER

ULTRASONIC

ALTIMETER

BAROMETER

GYROSCOPE

THERMAL

ELECTRICAL SPECIFICATION

Microcontroller: - Raspberry Pi 3B

GSM board: - 4g LTE

Battery: - LI Ion 12v 5 AH

ELECTRONIC SPEED CONTROL MODULE(Driver)

CAMERA SPECIFICATION (1/2.3" CMOS Huddly)

2x OPTICAL ZOOM

2 km 1080p VIDEO TRANSMISSION

Pixels: 12 MP

Burst Shooting (Resolution FHD)

1. 5. Strategies for implementation of social distance between two individuals: -
      1. STRATEGY NO: - 1
1. Recording will be live feed to dedicated police station.
2. In case of severe mass gathering buzzers and notifications will be given to police for acting.
3. Through Face recognition direct penalty can be charged also notifications can be given through mobile.
4. Shop owners can be warned and can be charged by local police.
5. In traffic vehicle number can be tracked for fining or informing.
6. Incase of emergency location and photos will be sended to local police station.

2. STRATEGY NO:- 2

For Industry and college)

(Unique Identity Code)

1. Industry generally have high ceiling which is enough for drone flight and refineries have open roof, college campus monitoring.
2. In this industry live feed will be directed to security office and employee will be directly penalized or informed through its employee code on its cloth
3. Compulsory mask and no spitting can be easily monitored in all areas.
4. Employee for maintain social distance should be rewarded.

1. 3. STRATEGY NO:- 3
1. It should constantly buzzer and should motivate people for maintaining social distance, without charging any penalty and live video feeding.
2. It should identify people with fever symptoms through temperature sensing and also people spitting at any corner.
3. It should bring awareness by playing audios in public market for social distance.
4. It should be capturing images at time of night curfew.

6. Algorithm

Figure 4 Flow chart of algorithm 1

1. Image analyzation algorithm

Figure 5 Image analyzation algorithm

• 2. Human detection
• Human detection is perform using the “ Arduino Object Detection Tracking”.
• Arduino object Detection Tracking is useful for the detection of the human.

4. Uniqueness

1. Smart Battery Charging Station: -

As drone has very less flight time it is very essential to charge it, but when we talk about surveillance system it should be autonomous for safety and reducing fatigue, to avoid it we programmed a smart battery charging algorithm which enables the drone to fly to its nearest charging station setup for charging, where there will be a subsidiary drone for its place in network. It will be a wireless pad which will charge the drone in 25 mins for 30 min of flight time.

2. Dynamo in drone: -

With increase in technology of battery-operated vehicle we found a small dynamo but of high capacity, the rpm of propeller will be 22000 and if we rotate dynamo at such a high speed it would generate decent amount of electricity. We have placed dynamo by nylon gear arrangement with propeller for increasing flight time.

Figure 6 Wireless Charging Pad

Figure 7 Dynamo with propeller

1. Result and discussion

COVID-19 proliferation/spread is best contained when the general public practices social distancing as a standard procedure in all public places no matter whether they are infected or not. Since the COVID-19 virus can survive on a variety of surfaces from hours to days, safety precautions with respect to wearing masks, gloves, sanitizing hands and maintaining personal hygiene is very important.

This system is to track them while raising alarms to concerned authorities by verifying their body temperature and whether they have worn masks & gloves. Subsequently, alerting individuals and authorities who might have come in contact with COVID-19 patients.

Hence, AI powered system identifies any unusual crowd movement, track group gatherings and also count people in real time at public areas using Drones. With the use surveillance cameras/Drones combined with thermal imaging, we can identify the following metrics in contactless way without endangering the health of enforcement personnel:

People violating any one or more of the metrics will be identified, tracked and their faces shall be saved in the databases and alerted to the concerned authorities in the premises for violation of social distancing and personal hygiene. The saved face images could be added to the “violator repository” and used to identify people if they are tested positive for COVID-19 later.

Based on tracks of the identified persons, disinfection of those areas can be performed to prevent any further chances of spread to healthy individuals. Also, all people who might have come in contact with the identified person will be notified to maintain caution and advised to practice stricter personal hygiene to prevent emergence of diseases. It is a move to make premises safer for people to move freely in the current situation where prevention of a community transfer is in progress. It is a major step towards enforcing social distancing on a large scale.

Hope it helpful to you for any doubt comment below i will solve your query
Please rate my answer your rating important to us.