Umpire’s Assistant : DRS with computer vision

Abstract

In any game, a fair decision is necessary to ensure that the game is played fairly. Any bad judgement made due to human error could determine the game's outcome. The literature study mentioned computer vision and image processing algorithms, which were shown using numerous cameras. With the use of a high-quality smartphone camera, this study focuses on a system that assists the umpire in making decisions such as no-ball, LBW (leg before wicket), Run out, stump out, and so on. The DRS (Decision Review System) is designed to make decisions such as run-out and stump-out. Tintern is used to create DRS's graphical user interface. The Histogram of Gradients (HOG) and Support Vector Machine (SVM) are used to classify and recognize objects (SVM).

Using the OpenCV library, we optimized and applied frame subtraction, contour detection, and minimal enclosing circle methods to detect the cricket ball from the video. To follow and forecast the ball's movement from a video source, linear regression and quadratic regression are utilized. For the visual representation, Python is used.