Classified Segmentation Method of Carbon Fiber Composites Based on Electrical Impedance Tomography

Carbon fiber reinforced polymer (CFRP) is widely used for its light weight and high ratio of weight to strength. However, the CFRP plate become defective or damaged due to the complex structure of CFRP. Electrical impedance tomography (EIT), an emerging method for monitoring damages in CFRP, has been widely investigated for its non-destructive, non-intrusive and low cost. The meshes for solving the inverse problem of EIT has many segmentation ways. In this paper, an classified segmentation method with L1 regularization (CS_L1) is proposed to make the details of damage more clear. The correlation coefficient (CC) is set as the evaluation index to estimate the similarity between the real damage distribution image and the reconstruction image. Both simulation and experiment are taken to prove the effectivity of this method.