Signal enhancement and imaging for air-coupled ultrasonic wood defect detection using hybrid mode decomposition

In air-coupled ultrasonic wood defect detection, received signals usually suffer from a low signal-to-noise ratio because of severe acoustic attenuation in air, low air-solid transmission efficiency, and noise contamination, which degrades image contrast and defect identification. To address this problem, this paper proposes a signal enhancement and imaging method based on hybrid mode decomposition, including signal denoising, feature extraction, and image denoising. In the denoising stage, variational mode decomposition (VMD) is first used to decompose the original signal, determine the optimal number of modes by information entropy, and reconstruct the signal with effective components selected by correlation coefficients. Empirical mode decomposition (EMD) is then applied for secondary denoising by removing spurious components. Simulation results show that the proposed method improves the average SNR by 14.18 dB and reduces the average MSE by 96.10%, outperforming conventional methods. In the imaging stage, time-frequency ridge energy peaks are extracted as imaging features, and the Total Variation (TV) algorithm is introduced to enhance image quality. Experimental results demonstrate that the proposed method can reliably identify wood knot defects and distinguish surface cracks from through cracks. Furthermore, evaluations of defect size accuracy and Weber contrast confirm its effectiveness and engineering applicability.