Automated Run out detection system in cricket using Deep Learning Techniques
Modern day cricket is achieving new heights and the popularity of this exciting sport is increasing day by day.Nowadays people are getting more and more enthusiastic about cricket and cricket comes next to football when it comes to viewership.The 2015 ICC Cricket World Cup is estimated to have been
2025-06-28 16:30:23 - Adil Khan
Automated Run out detection system in cricket using Deep Learning Techniques
Project Area of Specialization Artificial IntelligenceProject SummaryModern day cricket is achieving new heights and the popularity of this exciting sport is increasing day by day.Nowadays people are getting more and more enthusiastic about cricket and cricket comes next to football when it comes to viewership.The 2015 ICC Cricket World Cup is estimated to have been watched by an estimated 2.2 billion people.But every sport has its own flaws in terms of decision making which in this case, is the role ofthird umpire especially with the decisions involving run out and stumping. Because it is practically impossible for the umpire on field to give run out decisions with naked eye for which the decision is referred to the third umpire whomakes the final decision after playing the video of the run out in slow motion. The aforementioned process is a bit time consuming (which takes a minute or two) , involves human error and serves as a pause/break in the proceedings of the match and henceaffecting the rhythm of the batsmen and the bowling side as well. To eradicate this problem from the fast game of cricket we are proposing an intelligent system using Deep Neural Networks [DNN]for making instant run out decisions hence saving time byperforming the role of third umpire. Since the advent of DNN, Deep Learning models with their multi-level structures is very helpful in extracting complicated information from input images.DNN are also able to drastically reduce computation time by taking advantage of GPU for computation which many networks fail to utilize.Our design will primarily be acting as a classifier/ efficient decision maker, deciding between run out and not out decisions quickly.Our project will be starting with the collection of appropriate data on which feature extraction can then be performed so that Deep Neural Network architecture can be developed (Or trained).For the training of DNN intelligent system we will use Python programming language for which we will use PyTorch or Tensorflow etc. PyTorch or Tensorflow is a framework which is particularly developed for implementing Deep Learning architectures. After training the network, we will test our system through different sample data in order to determine its accuracy before being finalized.
Project Objectives- Automating the role of third umpire in decision making in cricket for providing a fast and accurate decision without any human error.
- The system designed can be further updated and scaled for other decisions like stumping, no ball etc.
- Cheap solution without changing conventional cricket setup
We are proposing an intelligent system using Deep Neural Networks [DNN] for making instant run out decisions hence saving time byperforming the role of third umpire. Since the advent of DNN, Deep Learning models with their multi-level structures is very helpful in extracting complicated information from input images.DNN are also able to drastically reduce computation time by taking advantage of GPU for computation which many networks fail to utilize.Our design will primarily be acting as a classifier/ efficient decision maker, deciding between run out and not out decisions quickly.Our project will be starting with the collection of appropriate data on which feature extraction can then be performed so that Deep Neural Network architecture can be developed (Or trained).For the training of DNN intelligent system we will use Python programming language for which we will use PyTorch or Tensorflow etc. PyTorch or Tensorflow is a framework which is particularly developed for implementing Deep Learning architectures. After training the network, we will test our system through different sample data in order to determine its accuracy before being finalized.
Benefits of the Project- Local Cricket (Domestic level) ; Central Punjab, Northern, Khyvber Pakhtunkhwa, Southern Punjab, Sindh, Balochistan.
- County Cricket (at International Level)
- International Cricket Council (ICC)
- Moin Khan Academy
- UBL Academy
- Customs Academy
- This system enables us to automate the run out decisions in cricket using Deep Neural Network.
- The system also saves the time in cricket.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 80000 | |||
| Canon EOS 1200D DSLR Camera with EF-S 18-55mm IS II Lens | Equipment | 1 | 30000 | 30000 |
| MSI Geforce GTX 1660 Armor 6G OC Video Graphics Card 6GB GDDR5 | Equipment | 1 | 32000 | 32000 |
| Camera accessories | Equipment | 1 | 3000 | 3000 |
| Cricket Kit | Equipment | 1 | 5000 | 5000 |
| Printing | Miscellaneous | 1 | 5000 | 5000 |
| Stationery | Miscellaneous | 1 | 5000 | 5000 |