Automatic Identification System of Pinewood Quality Level

The current forestry enterprises represent an important source of income for several communities worldwide. Pinewood transformation into a product is commonly manual, slow and costly. The high variability of the raw material is one of the challenges during the manufacturing process. For this reason, quality control is a key feature during the selection process, as all subsequent stages can be affected. In most cases, the sorting of pine wood is done manually, resulting in a biased, time-consuming, impractical and less effective process. Usually, pattern identification is carried out one by one by experts who inspect the wood surface texture. In this work, an automated system based on computer vision using OpenCV is proposed for sorting pine wood. To validate the proposed system's capability, a case study is presented. The sorting of pine wood is carried out according to the Mexican standard NMX-C-018-1986, which establishes six levels of quality. The results obtained show a high effectiveness of classification and a promising automatic solution for sorting pine wood.