Monitoring vertical density profile (VDP) of oriented strand board by computer vision

Objective The vertical density profile (VDP) is an important factor affecting the physical and mechanical properties of oriented strand board (OSB). Previous studies have mostly used indirect methods such as simulation prediction to obtain VDP. To quickly, accurately, directively obtain the VDP of OSB during the hot-pressing process, computer vision technology was applied in online monitoring of VDP.

Method Computer vision techniques was used to handle photographs collected during the hot-pressing process, to quickly and accurately identify and mark wood flakes, calculate the instantaneous compression ratio and density, and achieve online monitoring of VDP. Two types of OSB, with conventional and uniform VDP, were prepared in the laboratory. During hot-pressing process, the online technology for monitoring VDP was used to track and record the positions of the marked strands, and calculate the compression ratios and partial density. Thermistors were used to synchronously record the temperatures of surface and core layers of the board. Cross-sectional density scanner were used to calculate partial density. The accuracy of this technology was verified by comparing the actual results with the monitoring results.

Result (1) This technology was characterized by its speed, accuracy and completeness of data. (2) The mean correlation coefficient between instantaneous average compression ratio of the board and that of the marked strands was greater than 0.990. The correlation coefficient between the monitoring values and actual values of partial density was greater than 0.840. So, the online technology for monitoring VDP can accurately obtain the instantaneous compression ratio and partial density of the board. (3) The online technology can directely reflect the formation process of VDP during the hot-pressing process, providing a new method for interpreting the formation principles of VDP.

Conclusion A technology for online monitoring VDP of OSB in real-time based on computer vision has been developed. The application of computer vision can provide a new technical choice for optimizing the production of wood-based panels.