Case Study- Activity and Object Detection for Tennis

Wireless Camera Manufacturing Company

Activity and Object Detection for Tennis

Summary

With offices in North America, Europe, and Asia, this organization is the top manufacturer of wireless portable digital magnification cameras that provide real-time video transmission and works with resellers and dealers worldwide. Folio3 developed a state-of-the-art computer vision-based AI solution that processed high-resolution images and videos of the tennis ball in real-time to determine the ball's exact location in a game.

About the Customer

A leading provider of wireless portable digital magnifying cameras with real-time video transmission. The company’s product portfolio includes a wide variety of wired and wireless cameras and accessories. Its cutting-edge optics technology gives users crystal clear vision, pictures, and video.

Logo - frist secton (1)
  • Team composition

    5-6 members

  • Expertise used

    Object Detection, Object Tracking, Real-Time processing

  • Duration

    4-6 months

  • Services provided

    AI Image processing, Front-end Application, Back-end engineAI Model Development, Web Interface & Dashboards, AI Model Optimizations

  • Country

    US

  • Industry

    Wireless Camera Manufacturing

Understanding the Challenge

During a game, tennis balls could be tracked using portable magnifying cameras, but it was difficult to detect and track fouls and penalties. The company required a solution that could track a tennis ball in real-time, detect the exact location of the ball and determine if the ball is landing in specific pre-determined zones of the court.

Solution

Folio3 developed and trained the machine learning model to be deployed in the solution and also built a user interface to manage and run the solution with a descriptive reporting system.

Real-Time Detection

The solution utilizes machine learning to provide an accurate location of the ball in real-time during the game, using live video from HD digital cameras installed on the game field

real time detection

Front-End User Interface

A user interface was developed to show the results and points, every time the tennis ball landed off course or when there was a fault.

frontend user interface
white label solution
real time tracking

Real-Time Tracking  

IReal-time tracking of all ball and player movements to determine ball and player speed, size, and position

descriptive reporting system

Descriptive Reporting System

A descriptive reporting system that displays results in real-time after receiving data from trained AI models.

Result

With Folio3 AI’s solution, our client was able to detect and track the tennis ball with an accuracy of up to 90%, within an ongoing game in real-time, which automated the scorekeeping process, and defined fouls/faults and penalties with more precision thereby creating a smooth and delightful experience for players.