• MT24_Banner_1.png

Workshops

Workshop Title   :   Image Classification, Identification and Near Miss Detection Using You Only Look Once (YOLO) From CCTV Data
Speaker(s)   :  Assoc. Prof. Dr. Ahmad Sufril Azlan Mohamed and Dr. Majid Khan Majahar Ali
Duration   : 2 hours

  

Synopsis:

Computer vision-based methods have been widely employed for monitoring road traffic conditions. However, the judgment and prediction of near-miss incidents remain a challenging area that requires further research. Existing studies primarily rely on investigations, interviews, and police reports, which focus on areas where crashes or near-misses have occurred in the past. Consequently, this data is outdated and may not accurately reflect the current traffic situation.

Due to the lack of image data and research results on traffic near-miss incidents involving two or three-wheeled vehicles (such as motorcycles, bicycles, etc.), research in this domain is increasingly scarce. Nevertheless, there is significant potential for improving image recognition systems to obtain more accurate detection and tracking results for two or three-wheeled vehicles.

In this workshop, we will teach using YOLO deep leaning model, an improved object detection and tracking system that combines various methods, motion coefficients, and trajectory prediction mechanisms using our developed algorithm. This system will be trained on CCTV data collected from traffic scenarios. With Python, a coding language with a simple and easy-to-use syntax, is an excellent choice for individuals learning computer programming for the first time. 

Enroll in our course now to enhance your career prospects and acquire a valuable skill that will serve you well in your current and future endeavours.

This 2-hour workshop will cover the following topics:

  • Installation of Python-Ananconda
  • Python Basics
    1. Introduction to Jupyter Notebook
    2. Logical & arithmetic Operators
    3. Output Formatting
  • Main content
    1. Basic CCTV video
    2. YOLO deep learning model
    3. Image identification
    4. Vehicle Detection
    5. Near Miss Calculation

 

Objectives:

  • To demonstrate the use of YOLO deep learning model through Python programming language and using CCTV video in solving simple complex real-based problems using developed mathematical algorithm.

 

Who should attend: 

This workshop is suitable for anyone interested in learning YOLO deep learning model with Python programming for processing CCTV video.  

 

Requirements: 

Please bring your own laptop for this workshop. Your laptop should be running the 64-bit Windows operating system. Users should have administrator privileges on the laptop for software installation.