A Customisable AI Deck for Pitch Reports and Automated III Umpire Decision Review System DRS

Abstract

Nowadays giving fair verdict is a quite challenging task because of certain contentious aspects in modern cricket. So, in order to avoid making wrong decisions, we develop an automated AI-based solution. This project focus on a technology that helps both the main umpire and third umpire to makes critical determination for Leg Before the Wicket (LBW) regarding whether the batsman is out or not-out and also minimizes the waiting time for players until the third umpire go through the trajectory of the ball to make a correct decision. The main purpose of our AI-DRS is to remove the umpires call which plays a vital role in giving third umpires decision because whether any one of the cases shows umpires call the decision will be stick with on-field umpires call whether it may be out or not-out. The pitch report and comprehensive cricket laws are also included for the sake of the game. The pitch report will be examined with several key wicket characteristics, such as kind of soil, cracks, amount of grass cover, and wetness, etc. using drone we capture the video of the match day pitch. To determine the field crack, canny edge detection is performed and soil moisture sensor is used to determine the moisture content of the soil. This information help cricket team to make a decision about whether to bat or field after winning the toss and helps to choose the strongest 11 players through which can win the match on that pitch on that day. Utilizing support vector machine (SVM) and histograms of gradients (HOG), objects are classified and recognized. In order to monitor and forecast the velocity of the ball, linear regression and quadratic regression are applied. Finally, Tkinter is used for GUI development, imutils and OpenCV are used as implementation tools. Due to the controversy of rare wicket calls, boundary and penalty runs, we bring a voice recognized AI system which gave fans to easily understand why this decision is made by the umpire and sometime umpires found difficulty to remember some rules which is rarely used in cricket it will also give assist to on-field umpires to give a very clear idea why he made the decision, the on-field umpires can easily access the laws through voice recognition which use Alan-AI. The Voice recognition web app was developed using react-js.