Welcome to visit Scientia Silvae Sinicae,Today is

Scientia Silvae Sinicae ›› 2021, Vol. 57 ›› Issue (9): 181-192.doi: 10.11707/j.1001-7488.20210918

• Scientific notes • Previous Articles    

Stress Wave Tomography Imaging Algorithm Based on Ray Segmentation for Nondestructive Testing of Wood

Tao Liu,Guanghui Li*   

  1. School of Artificial Intelligence and Computer Science, Jiangnan University Wuxi 214122
  • Received:2020-05-06 Online:2021-09-25 Published:2021-11-29
  • Contact: Guanghui Li

Abstract:

Objective: Stress wave nondestructive testing technology has been widely applied in wood and tree defect detection. The traditional stress wave imaging algorithm has low accuracy due to the limited sensor number. To improve the accuracy of the stress wave imaging algorithm and accurately reflect the position, size and decay of the internal defects of wood, the signal distribution of the stress wave on the cross section was investigated. Method: Four log samples(Pinus spp., Sapium sebiferum) and four live trees(Cinnamomum camphora, Sabina chinensis, Salix spp.) in Jiangnan University campus were selected as the samples for experiments, and the stress wave velocities were collected using FAKOPP instrument. Firstly, the stress wave velocities were corrected, and the wave ray diagram was generated, then the cross section was divided into many grid cells. Secondly, each stress wave ray was segmented, and the velocity of multiple segments on the ray was estimated, therefore the stress wave signal amount was increased and an improved stress wave ray diagram was also obtained. The velocity values of the grid cells were estimated using the improved propagation rays and the image processing method, thus the tomographic image of wood was generated. The Resistograph micro-drilling tool was used to evaluate the internal conditions of the live trees. Result: Experimental results showed that the proposed algorithm could accurately reconstruct the tomographic images of the four log samples. The Resistograph micro-drilling tool was used to perform multi-path drilling on live trees to get the resistance curves. Compared with the resistance curves, it was demonstrated that the proposed algorithm could also generated the tomographic images with a high resolution of 4 live trees. Conclusion: Based on the ray segmentation, the proposed algorithm might improve the accuracy and correlation of the initial grid cell velocity value, which could achieve high-quality imaging of wood defects, and could be effectively applied to nondestructive testing of wood.

Key words: stress wave, non-destructive testing, ray segmentation, acoustic tomography

CLC Number: