ECE 492 Computer Engineering Design Project Curling Coach

  • Slides: 1
Download presentation
ECE 492 - Computer Engineering Design Project Curling Coach Nicole Stodola, Gerry Finlay, and

ECE 492 - Computer Engineering Design Project Curling Coach Nicole Stodola, Gerry Finlay, and Chris Pederson Project Motivation • • • 40+ years of curling experience between Nicole and Chris Wanted to improve the current delivery verification method Level shoulders is very important to the curling delivery 2013 Project Design • Nikon Coolpix P 5000 digital camera captures the video of the curler • Image is displayed onto the VGA monitor with red boxes marking the targets • Image processing module detects if there is a vertical deviation in the shoulder targets • If the deviation is above the 5 pixel threshold an alarm is triggered • The system alarm, a short audio clip, is played and green LEDs are turned on Components • Altera DE 2 FPGA board • Digital camera with NTSC video capability • VGA Monitor • Speakers Project Overview Testing § Video processing of a curling delivery VHDL Testing § Provide a method of instant notification of misaligned shoulders • Test each module separately • Camera § Ability to add in more target body parts for tracking • Audio • Image Processing § Video display with markings showing target positions and feedback • Test benches for all VHDL System Testing Fig 2: System in action with athlete Image Processing • Video frame is split into two sections for target detection • In each section the system scans from the top left for the greenest pixels • Once the pixels are detected it draws a red box on the video as shown in Figure 2 Fig 1: Athlete sliding out • The system tracks X and Y coordinates of the targets with red boxes and monitors the Y value for a predefined deviation in vertical position • Once a deviation in the targets has been detected, trigger the alarm condition • Testing in the lab • Testing in a curling rink • Edge case considerations: • Different slides • Obscure targets Group 15 would like to thank Dr. Duncan Elliott, Nancy Minderman, Michael Shoniker and Jessica Cao. Department of Electrical & Computer Engineering