AI Base Thundercam using Raspberry Pi and IoT

Raspberry Pi and Ai integration

1. AI BASED THUNDERCAM USING IOT AND RASPBERRY PI

2. GROUP MEMBERS STUDENT NAME CMS ID MUHAMMAD WALEED KHAN 45342 SHERAZ AHMED 45260 ABDUL TAHIR KHAN 45804 SUPERVISOR SUPERVISOR NAME DEPARTMENT Supervisor SIR MEHRULLAH SOOMRO Computer Engineering 2 Department of Computer Engineering Ai based ThunderCam using IoT and Raspberry Pi

3. Motivation Introduction Literature Review Problem Statement Project/Research Objectives Proposed Methodology Project Work Load Gantt Chart References Outline 3 Department of Computer Engineering

4. Motivation 4 The main Motivation occurred when we decided to Utilize the fields we have previously studied and explored like Artificial Intelligence and Web Development which eventually acceded all the group members on the highlight to make a economical friendly, Pakistan's made AI network CCTV camera. Department of Computer Engineering

5. Introduction 5 Department of Computer Engineering CCTV Control unit Software Hardware Interfacing of Hardware Detection using Open CV Performance/Testing Testing in different Mock Environments Decision making Unit: -Generating Notification Training Data using Tensor Flow

6. Literature Review 6 Department of Computer Engineering Name of the Paper Year Published Pros & Cons Summary Low cost smart security camera with night vision capability using Raspberry Pi and OpenCV [6]. 2014 - Only applicable when Human Detection takes place. - To optimize computation Background Subtraction is used. Specifically made for Night Vision using Infrared Technology An Internet of Things Approach for Motion Detection using Raspberry PI [7]. 2015 - Attributes can be altered accordingly. - RPI Model can’t run x86 Operating Systems Motion detection is done using Python script. An IOT Approach for Motion Detection Using Raspberry PI [8]. 2017 - Real-time notification is generated. - It requires Internet Connection and also can’t run on x86 Systems Motion detection done using Python

7. Literature Review 7 Department of Computer Engineering Name of the Paper Year Published Pros & Cons Summary Smart Surveillance Monitoring System [9]. 2018 - System uses a 3G Dongle to sent video transmission to Device. - Frame Rates aren’t controlled precisely. PIR sensor is used for Monitoring purpose A vision-based home security system using OpenCV on Raspberry Pi 3 [10]. 2019 - RPI camera module is used to minimize computations. - Security for SMTP isn’t good Face and Motion Detection is done suing Open CV Packages. Development of portable automatic number plate recognition system on Raspberry Pi [4]. 2019 - An adequate system is developed using proper Image Processing Techniques System uses portable ANPR that runs on Raspberry Pi

8. Problem Statement 8 The modern era for security purposes as we can see is shifting on AI-based Smart Surveillance Cameras, but on the Contrary, importing such smart CCTV cameras put out a heavy burden on the nation. Department of Computer Engineering

9. Objectives 9 Preparing Raspberry Pi for Project Streaming CCTV live stream to Raspberry Pi. Training Taffic Rules and Robbery Single Model Converting installed CCTV into ThunderCam Department of Computer Engineering

10. SDG’s 10 Our Project will be accomplishing 2 vital SDG’s, which are as following: SDG-11 Make cities and human settlements inclusive, safe, resilient and sustainable SDG-16 Promote peaceful and inclusive societies for sustainable development, provide access to justice for all Department of Computer Engineering

11. Objectives (Cont.) 11 Department of Computer Engineering OBJECTIVE-01 1 OBJECTIVE-02 2 OBJECTIVE-03 3 OBJECTIVE-04 4 Commencing and Setting-up Raspberry Pi environment Getting Live-Stream from CCTV through Raspberry Pi Training and availing the datasets of Traffic Rules Violation and Robbery. Utilizing the installed cameras by connecting this very system Current Progress:

12. Proposed Methodology 12 Department of Computer Engineering Tools used: Hardware Tools Raspberry Pi CCTV Camera Software Tools Open CV TensorFlow Lite LabelImg Google Colab Gotify

13. Proposed Methodology 13 Department of Computer Engineering Traffic Rule Violation Pursue with Live Streaming END Yes Yes No Yes No No Start Process Video Image Acquisition Is Motion Detected? To Identify Potential objects Rider using phone or w/o helmet ? Take actions: Generate Notification

14. Proposed Methodology 14 Department of Computer Engineering Robbery Detection Pursue with Live Streaming END Yes Yes No Yes No No Start Process Video Image Acquisition Is Motion Detected? To Identify Potential objects Robbery In progress? Take actions: Generate Notification

15. Proposed Methodology 15 Department of Computer Engineering Initial Stage of the project Finding the problems Creating the Rough Sketch Estimating the Components and Cost Testing the Components Installation of Raspbian Interfacing the CCTV Interfacing Programming Testing Getting Notifications Configuring Gotify with RPi Installation of Packages Installation of OpenCV/Tensor Flow Lite Gathering the Data-set Training of Model

16. Results 16 Department of Computer Engineering Installation of Raspberry Pi

17. Results 17 Department of Computer Engineering Live-Streaming through website & Raspberry Pi

18. Results 18 Department of Computer Engineering Motion Detection

19. Results 19 Department of Computer Engineering Custom Trained Model Results

20. Project Workload 20 GROUP MEMBERS CMS ID TASKS DISTRUBION Abdul Tahir Khan 45804 Hardware Interfacing. Sheraz Ahmed 45260 Raspberry Pi Programming & Web Development. Muhammad Waleed Khan 45342 Literature Review, Training Dataset & Documentation. Department of Computer Engineering

21. Gantt Chart 21 Department of Computer Engineering

22. References 22 Department of Computer Engineering [1] J. Blanes I Vidal et al. , “The Effect of Police Response Time on Crime Clearance Rates *,” 2017. [2] “SMART SURVEILLANCE MONITORING SYSTEM,” 2018. [Online]. Available: www.worldwidejournals.com [3] P. B. Patel, M. T. Student, V. M. Choksi, S. Jadhav, and M. B. Potdar, “ISSN : 2249-0868 Foundation of Computer Science FCS,” 2016. [Online].Available: www.ijais.org [4] S. Fakhar A. G., M. Saad H., A. Fauzan K., R. Affendi H., M. Aidil A, “Development of portable automatic number plate recognition (ANPR) system on Raspberry Pi2015, doi: 10.17148/IJARCCE.2015.4749. [5] G. Nagalakshmi, D. I. Bhargavi, G. Nandini, and P. G. K. Sirisha, “REAL TIME CRIMINAL DETECTING SYSTEM Using Raspberry pi,”

23. References 23 Department of Computer Engineering [6] Wilson Feipeng Abaya; Jimmy Basa; Michael Sy; Alexander C. Abad “Low cost smart security camera with night vision capability using Raspberry Pi and OpenCV” [7] Aamir Nizam Ansari Mohamed Sedkyl , Neelam Sharma2 , Anurag Tyagi “An Internet of Things Approach or Motion Detection using Raspberry Pi” [8] Neha Patil; Shrikant Ambatkar; Sandeep Kakd “IoT based smart surveillance security system using raspberry Pi”. [9] Maya Nayak, Prasannajit Dash, “SMART SURVEILLANCE MONITORING SYSTEM”. [10] Thinesh Prathaban, Weilynn Thean and Mohd Ilyas Sobirin Mohd Sazali “A vision-based home security system using OpenCV on Raspberry Pi 3”.

24. 24 Thank You Department of Computer Engineering