Abstract: In order to improve the accuracy of wood-based panel thickness detection, a wood-based panel thickness detection method based on improved Harris Hawk Optimization（HHO） algorithm and Extreme Learning Machine（ELM） was proposed. By improving the HHO algorithm, and using the optimized algorithm to select parameters such as weight and bias values of extreme learning machine, the optimization mechanism was retained on the basis of improving the performance of the algorithm. At the same time, reverse learning of tent mapping was introduced into the initial population position to reduce unnecessary global search and improve the convergence speed of the algorithm without affecting the population diversity. Finally, taking Medium Density Fiberboard（MDF） as an example, the on-line detection experiment was carried out, and the experimental data were obtained and compared. The experimental results showed that the proposed method can effectively reduce the measurement error and improve the measurement accuracy, which had a certain practical application value.