Playing%20Chess:%20Recognition%20and%20Simulation%20COMP%20290-075%20Computer%20Vision

1
Playing Chess Recognition and SimulationCOMP
290-075Computer Vision

• B. Danette Allen
• Paul J. McLaurin
• May 03, 2000

2
Objectives

• Track individual chess pieces
• Maintain state of board
• Graphically represent state changes and state

3
Lab Set-up

• 1 to 3 cameras
• Panasonic GP-LM7TA
• 1 overhead (min)
• 2 side view
• PC
• Matrox imaging board and libraries
• C, OpenGL, GLVU
• Chessboard and pieces

4
System Constraints

• Single move per image
• move, capture, en passant, castle
• Shadow minimization
• currently using halogen lamp (no overhead)
• Not constrained to legal chess moves
• Must start game from beginning

5
Calibration of chess board
6
Blob Analysis

• A blob is a set of connected pixels that have the
  same intensity value
• There are five steps in this process
• 1. Acquire an image
• 2. Analyze foreground and background pixels to
  determine optimum threshold required to segment
  objects from their background
• 3. Apply a threshold to create a binary image
  with objects shown in white and background shown
  in black
• 4. Analyze areas of connected white pixels and
  assign a label or number to each discrete group
  of connected pixels
• 5. Extract physical measurements from objects.

7
Grayscale and Binarized Images
8
Rendered chessboard and pieces
9
Problems Encountered

• Black on black not detectable
• white ring around base of black pieces
• black ring around bases of white pieces
• Multiple responses
• Added code to filter out redundant responses
• Noisy difference images
• Gaussian filter

10
Future Enhancements

• Multiple/Side cameras
• Tracking of individual pieces
• discrete tracking of pose
• continuous tracking (6 DOF)
• Handle shadows
• Increase code robustness
• Verify legal moves in all cases
• Handle nudged pieces automatically

11
Demonstrations

• Video of chess game in lab
• Simulated chess game