The log end scanner is an additional module for the VECTOR-3D measuring device improving the accuracy of log accounting and sorting due to a qualitative transition from "geometric" to "analytical" vision. While a 3D laser scanner sees the log's external shape, the end scanner sees its actual content and the actual condition of its cross-sections.
A pair of video modules is installed on the VECTOR-3D frame, photographing the ends of the sorted logs. Artificial intelligence determines the actual diameter of the butt and top without bark. The system compares this with laser scanning data and automatically accumulates statistics by species, batches, and specific suppliers. Unlike traditional methods, our solution does not rely on tabular coefficients. Therefore, seasonal changes in moisture content and the characteristics of the wood species do not affect the measurements.
Based on the obtained data, the log taper and bark coefficients are also calculated. Additionally, the trained neural network identifies key wood defects in the log end: rot and blue stain spots measuring 30x30 mm and larger, and fissures over 2 mm wide.
All this increases sorting accuracy and reduces the number of out-of-order logs, thereby increasing the overall yield of lumber. On average, the neural network provides a 0.6% volume correction for a batch.