Wood Surface Inspection
Identify quality problems in final cut wood boards
Graphical programming environment for deep learning-based industrial image analysis
Modern sawmills or lumber mills are computer-controlled enterprises that process raw logs into millions of board feet of lumber per day. After initial cutting and trimming, wood is kiln or air dried, then planed and cut to final dimensions. Wood is precisely graded for quality. The number and location of defects, whether chips, cracks, or other flaws, in the final cut determine the lumber grade, and thus the final price that can be charged.
As a natural product, wood is more varied than artificial materials, making wood surface inspection challenging. It is impossible to program all possible defects and wood patterns, so conventional machine vision is of limited use for lumber grading visual inspection.
AI-based vision systems and software are trained on images of various possible wood defects revealed by saw cuts. The classification tool then identifies and distinguishes various types of wood defects, while accepting the wide range of patterns, textures, and color variations.
Wood quality problems, whether a poor cut or poor quality of initial wood indicates a problem upstream in the supply chain. AI-enabled technology is sensitive enough to detect slight variations in the cut that indicate a possible problem with the saw blade. Instead of waiting for that saw’s quality to decline to the point that its cut is ragged enough to require a wood regrade, it can be adjusted or replaced long before there is any visible problem.