AI Bases Ball Tracking System
This project aims to develop a system with the help of camera?s, integration of microphone?s and various other sensors that assists the umpire to make decisions such as determining no balls, wide balls, bumpers, and most importantly leg before wicket (LBW). It includes the use of AI and Machine lear
2025-06-28 16:25:03 - Adil Khan
AI Bases Ball Tracking System
Project Area of Specialization Artificial IntelligenceProject SummaryThis project aims to develop a system with the help of camera’s, integration of microphone’s and various other sensors that assists the umpire to make decisions such as determining no balls, wide balls, bumpers, and most importantly leg before wicket (LBW). It includes the use of AI and Machine learning, Spin Line equations along with Computer vision algorithms for object detection and image processing. As manual decision is less reliable, so by automating the decision making process, it will automatically reduce the chances of human error and make the game of cricket more fair and efficient. The project seeks to create a Ball Tracking System which will create a virtual path for the Ball by tracking and calculating its line length, behavior, nature, and other natural parameters through various algorithms. In cricket a major issue in LBW decision is to determine whether it was Leg by or not. It is especially difficult for the naked eye to make an unbiased decision quickly therefore often times the umpire needs assistance to make the call. The UK-based company Hawkeye currently dominates the ball tracking system services throughout the cricket industry. Our group aims to make this service available domestically while making it affordable and improving the ball tracking system to better suit the domestic needs.
Project ObjectivesA system that can detect the movement of ball ,bat and person standing on pitch.
After detection, system must be able to create a trajectory path of ball, a virtual path showing its possible movements.
System must be able to store data that has been captured for future utilization and help.
Approach to create most economic system that can be availed by different complxes for inter matches then leading to higher oppurtunities.
Our basic goal is to create this technology so it can be implemented in university as model base for our start up.
We are working on this project to give a boost to our sports industry and to utilize our own resources for productivity.
Project Implementation Method- Study of Machine Learning and Deep Learning
- Histogram of Gradients (HOG), Sliding Window and segmentation by color are used for object classification and recognition.
- For, Frame subtraction, minimum enclosing circle, and contour detection algorithms are optimized and used for the detection of a cricket ball various algorithms are applied using the Open Source Python Library – Open CV.
- Data set is created on our own of various type so system can easily detect the things.
- An Algorithm Yolo V5 will be used for object detection.
- Use of Tensor Flow, for image recognition and processing.
- Machine Learning techniques - Linear and Quadratic Regression are used to track and predict the motion of the ball. It also involves the use of open source Python library VPython for the visual representation of the results.
- Economic system for services.
- Getting it to international platforms after recognition.
- To help sports industry.
- This tracking system can be used for other sports than cricket, eg Football, Bedminton, Basketball.
- It will effect pakistan's economy in future.
- Sports complex, Universities, National matches Management can avail services anytime if required.
- High Knowledge required, meeting with experience persons having vast knowledge in deep learnig.
- People having hardware knowldge with AI.
- Help from online platforms
- Updated softwares and Algorithms.
- Result accuracy.
- Prediction depends on system quality.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 70000 | |||
| Camera | Equipment | 3 | 15000 | 45000 |
| Software | Equipment | 2 | 5000 | 10000 |
| API | Equipment | 3 | 2000 | 6000 |
| Cricket item's | Equipment | 4 | 1000 | 4000 |
| Other | Miscellaneous | 5 | 1000 | 5000 |