An Intelligent Web Platform For Officiating in Cricket Sport
Donthireddy, Chandrasai Reddy
Citations
Abstract
While international cricket uses elite technology like the Decision Review System (DRS), regional and grassroots leagues are forced to accept game-changing officiating errors with no way to challenge them. This research was born from the frustration of losing matches to decisions that could not be reviewed. To address this, we developed an intelligent, low-cost umpiring platform designed to run on consumer hardware. The system uses a multi-modal AI stack combining YOLOv8 for object detection and MediaPipe for pose estimation to turn raw match footage into objective rulings, complemented by a custom Digital Signal Processing pipeline built on librosa and scipy for acoustic edge detection. The platform renders pose landmarks and ball-trajectory overlays onto decision video, giving umpires a transparent, frame-accurate view of every call. Field trials conducted in North Georgia demonstrate that the platform can deliver structured, evidence-based officiating on consumer hardware, bringing DRS-style fairness to levels of the game that have historically gone without it.
