Automatic Tracking of Traffic Violations Using Machine Learning

A final year project presentation on Automatic Tracking of Traffic Violations Using Machine Learning

1. ATRIA INSTITUTE OF TECHNOLOGY BANGALORE-54, Karnataka DEPARTMENT OF INFORMATION SCIENCE & ENGINEERING Project Phase -02 Project Review -01 “ Automatic Tracking of Traffic Violations ” VISVESVARAYA TECHNOLOGICAL UNIVERSITY “Jnana Sangama”, Belagavi– 590018 PRESENTED BY Pooja R(1AT19IS067) Harshita B M(1AT19IS035) Sankeerthi G K(1AT19IS091) UNDER THE GUIDANCE OF Mrs. Sonia S B Asst Professor Dept. of ISE, Atria IT

2. Automatic Tracking of Traffic Violations using Machine Learning

3. Project Exhibition Project Exhibition Name : TECHNOVA Project Exhibition attending Date: 05/05/2023 Project Exhibition attended College/University: RR Institute of Technology Include Project Exhibition attended Certificate(one sample Certificate)

4. Paper Publication Details Paper Title: Enforcing Road Safety: Machine Learning-based Automatic Traffic Violation Detection and Tracking. Journal/Conference name: INTERNATIONAL JOURNAL OF CURRENT SCIENCE - IJCSPUB (IJCSPUB.ORG) Journal ISSN: 2250-1770 Journal Vol : 13 Journal website: https://ijcspub.org/

5. CONTENTS Introduction Literature Survey Proposed System & Advantages Objectives Methodology/Algorithms System Design System Requirements Implementation with modules Result and Discussion References

6. INTRODUCTION: In the previous hardly any decades, noteworthy activities in the field of moving object recognition and following have been done to make following applications dependable, powerful and proficient: video observation, mechanical technology, verification framework, media creation, natural research and so on. There are numerous difficulties which produce leaps in the improvement of these applications. To avoid the difficulties traffic police must be accessible out and needs to constantly make sure if some vehicle is disregarding the standards. In the proposed framework, an answer for signal breaking infringement is given. These highlights are coordinated against predefined set of same vehicle number plate pictures in the database

7. The framework incorporates a mechanized framework by utilizing IR sensor, camera and number plate acknowledgment application. In the framework IR sensor will be set close to zebra crossing line. In the event that any vehicle crosses the zebra line, the work area application will be started and will catch number plate picture. Number plate acknowledgment application by utilizing picture handling calculation will perceive number plate and SMS will be sent to the guilty party if there should arise an occurrence of rule infringement scenario. In almost all the nations they provide guidelines for driving which are accessible to all residents to drive cautiously. The customary OCR based methodology for number plate acknowledgment doesn't work for the varieties in painting style of the number plates. In the proposed strategy an advanced mobile phone is utilized to catch the pictures and concentrate highlights of the vehicle number plate.

8. LITERATURE SURVEY: Sl . No Paper Title Authors, Publisher & Publication Year Problem Identified Techniques used Outcome 1 Intelligent Traffic Violation Detection system. Roopa Ravish , Kausthub Charon ,07 October 2021. The violations detected are vehicles jumping red signals. To validate this theory the author has utilized two techniques i.e. temporal violation and appearance-based violation. Traffic violations can be detected in a day light. 2 Traffic Enforcement System. Angel mariya T.P,Aishwarya M.J, March 2017. To detect whether the person is wearing helmet or not. It is a traffic management system using RFID. The system can track vehicles effectively, and automatically save and display the information. 3 Traffic signal violation detection Dr. S. Raj Anand , Dr. Naveen Kilari ,December 2021 Signal jump violation detection. PIC micro controller ,ML and Deep learning Results show that the detection of multiple traffic violations from a single input source is achievable.

9. 4 Prediction of Traffic-Violation Md Amiruzzaman ,2017 To detect traffic violation ,no number plate. Data mining Results show that the detection of traffic violation. 5 International analysis on social and personal determinants of traffic violations and accidents with elastic net regularization Yasuhiro Shiomi a,Azusa Torium , December 2021 Detection of triple riding. Elastic Net Regression model The results of the models for predicting the experience of traffic violations in the past two years. 6 Aanlysis of the impact of traffic violation monitoring on the vehicle speed of urban main road. Fuquan pan,Yongzheng Yang,March 2020 Analyze parking area detection. Image processing and communication. The results of the model is to analyze parking area detection.

10. 7 Video-based Traffic Violation Detection System Biaobiao Zhang,Ke _-Lin- Du,February 2015 It can detect traffic violations, such as running red lights. C++ with open CV,Block Match Algorithm. The results of the models for detecting traffic violations. 8 Traffic management system Allan N de souza,Roberto S Yokoyama,October 2016 It can detect number of person on vehicle. RFID technology. The results of the models for detecting number of people. 9 Advanced Traffic Violation Penalty System Mr.Pranob K Charles, Srilekha Govvala,june 2022. It can detect red light jumping,number of persons,no helmet detection. GSM , OCR Detects the traffic violation. 10 Detecting and Handling Traffic Violation M.Yogavalli , E.Arulmozhi , October 2015 It can detect jumping signal. RFID,PIC. Detects Traffic violation.

11. PROPOSED SYSTEM & ADVANTAGES: LPR is a ITS (Intelligent Transport System) innovation that not just perceives and checks the quantity of vehicles yet in addition separates them. For certain applications, for example, electronic cost assortment and red- light infringement implementation, LPR records tags alphanumerically so the vehicle proprietor can be surveyed the suitable measure of fine. In others cases, similar to business vehicle activities, a vehicle's tag is thought about against a database of adequate ones to decide if a truck can sidestep a checkpoint or a vehicle can enter a gated network or parking garage.

12. A tag is the exceptional recognizable proof of a vehicle. The essential issues in realtime tag acknowledgment are the precision and the acknowledgment speed. Tag Recognition has been applied in various applications, for example, naturally distinguishing vehicles in parking garages, get to control in a limited zone and recognizing and confirming taken vehicles. Nature of calculations utilized in a tag identifier decides the speed and exactness of the tag location. Before, various methods have been proposed for finding the plate through visual picture preparing. The System is used to detect and read number plates automatically 24/7 in real time. Installed model will help travelers to obey traffic rules Can operate simultaneously in multiple lanes Model detects and recognizes different dimensions, contrast , colors number plates with variety of character font and style.

13. OBJECTIVES: To replace the human traffic police by a Virtual machine To detect whether the motor cyclist is wearing the helmet or not To detect no parking rules To detect triple riding on the bike To detect the signal jumping To detect the vehicle number plate from the caught picture To send message to the respective vehicle owner

14. METHODOLOGY/ALGORITHMS: 1) Read Image from Webcam Capture image from webcam. Store up the caught image into a picture document for additional preparing. 2) Alter image to dim scale Convert picture into dim scale. Calculate suitable limit an incentive for the picture. Convert the picture into determined limit.

15. 3) Detecting Number plate zone Load tiny gaps including quantities of Number plate so number plate territory will be huge to seclude from figure. Determine width and tallness of the picture. Scan each and every pixel of line including number of white pixels in the accompanying framework. Crop the necessary zone. 4) Segmentation Channel the commotion level present in the picture. Separate each character from the plate.

16. 5) Number Distinguishing Proof Create the layout record from the put away format pictures. Compare each character with the layouts. Store the best coordinated character. 6) Save to record in given arrangement Open a content record in compose mode. Store the character acquired from the number distinguishing proof procedure to content document in given arrangement. Close the document

17. SYSTEM DESIGN: State Diagram

18. Architecture

19. Flowchart

20. SYSTEM REQUIREMENTS: Software Requirements Operating System : Windows 10. Software Tool : OpenCV,YOLOv5, OCR Coding Language : Python 3.8 Hardware Requirements PC : Minimum of 4/8GB Ram Hard Disk : 512 GB System : intel i3/i5 2.4GHz

21. IMPLEMENTATION MAKESENSE AI GOOGLE COLAB

22. YOLOV5 PYTHON IDLE TWILIO PYTORCH CUDA WEBCAMERA

23. RESULT & DISCUSSION 1.NO HELMET DETECTION IMAGES OUTPUT

24.   2.Triple Riding or More Than Two People Violation IMAGES OUTPUT

25. 3.No Parking Violation IMAGES Output

26. 4.SIGNAL JUMPING DETECTION IMAGES OUTPUT

27. REFERENCES: [1 ] S. Asoba , S. Supekar , T. Tonde and J. A. Siddiqui , "Advanced Traffic Violation Control and Penalty System using IoT and Image Processing Techniques," 2020 2nd International Conference on Innovative Mechanisms for Industry Applications (ICIMIA), 2020, pp. 554-558, doi : 10.1109/ICIMIA48430.2020.9074949. [2] N. Sulaiman , S. N. H. M. Jalani , M. Mustafa and K. Hawari , "Development of automatic vehicle plate detection system," 2017 IEEE 3rd International Conference on System Engineering and Technology, 2013, pp. 130-135, doi : 10.1109/ICSEngT.2013.6650157. [3] K. Yogheedha , A. S. A. Nasir , H. Jaafar and S. M. Mamduh , "Automatic Vehicle License Plate Recognition System Based on Image Processing and Template Matching Approach," 2018 International Conference on Computational Approach in Smart Systems Design and Applications (ICASSDA), 2018, pp. 1-8, doi : 10.1109/ICASSDA.2018.8477639. [4] M. Xu , H. Wang, S. Yang and R. Li, "Mask wearing detection method based on SSD-Mask algorithm," 2020 International Conference on Computer Science and Management Technology (ICCSMT), 2020, pp. 138-143, doi : 10.1109/ICCSMT51754.2020.00034. [5] S. Kadam , R. Hirve , N. Kawle and P. Shah, "Automatic Detection of Bikers with No Helmet and Number Plate Detection," 2021 12th International Conference on Computing Communication and Networking Technologies (ICCCNT), 2021, pp. 1-5, doi : 10.1109/ICCCNT51525.2021.9579898.

28. [6] Ahmed, N. J., Alrawili , A. S. and Alkhawaja , F. Z. (2020) “The Anxiety and Stress of the Public during the Spread of Novel Coronavirus (COVID-19)”, Journal of Pharmaceutical Research International, 32(7), pp. 54-59. doi : 10.9734/ jpri /2020/v32i730460. [7] Obed Appiah, Ebenezer Quayson , Eric Opoku, Ultrasonic sensor based traffic information acquisition system; a cheaper alternative for ITS application in developing countries, Scientific African, Volume 9, 2020, e00487, ISSN 2468- 2276,https://doi.org/10.1016/j.sciaf.2020.e00487. [8] R. Shreyas , B. V. P. Kumar, H. B. Adithya , B. Padmaja and M. P. Sunil, "Dynamic traffic rule violation monitoring system using automatic number plate recognition with SMS feedback," 2017 2nd International Conference on Telecommunication and Networks (TEL-NET), 2017, pp. 1-5, doi : 10.1109/TEL-NET.2017.8343528. [9] W. Zhang, C. -f. Yang, F. Jiang, X. -z. Gao and X. Zhang, "Safety Helmet Wearing Detection Based on Image Processing and Deep Learning," 2020 International Conference on Communications, Information System and Computer Engineering (CISCE), 2020, pp. 343-347, doi : 10.1109/CISCE50729.2020.00076. [10]Yuan Jing, B. Youssefi , M. Mirhassani and R. Muscedere , "An efficient FPGA implementation of Optical Character Recognition for License Plate Recognition," 2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering (CCECE), 2017, pp. 1-4, doi : 10.1109/CCECE.2017.7946734

29. THANK YOU