Digitizing Wood | Analyzing Wood Grain in 2x4s using Facial Recognition Software Strategies

Zippered Wood is a novel take on wood joinery and deformation in which digitally generated formally specific joint patterns are cut into boards that are joined to produce predictably precise bends. Within this system research is being done to maximise the usability of the 2×4’s and the strength of the zippered wood. By using waste stream 2×4’s for zippered wood we inherit problems that are not common in their traditional use. Namely those are the knots and screw/nail holes in the wood. Once the pieces are milled these areas create weak points in the veneer. Through our testing we have found that these areas are prone to cracking during the gluing process and create potential failure points along the piece. In response to this we have begun to develop a simulation algorithm to strategically map the zipper tooth surface within the 2×4 to find the optimal placement in relation to any de¬fects identified in the grain of the 2×4. This process, Digitizing Wood, uses an image of the 2×4 to locate the knots and defects in the wood to determine the optimal location. Using standard facial recognition techniques and imaging processing algorithms the knots and defects are marked. The marked regions and the zipper tooth surface are run through an evolutionary solver to optimize the placement in the 2×4. Once complete the finished piece is stronger and more materially efficient in its use. The Digitizing Wood strategy seeks to improve the effectiveness of the Zippered Wood system.

https://doi.org/10.35483/ACSA.AM.108.50

Volume Editors

ISBN
978-1-944214-26-5