EIT Image Processing Based on 2-D Empirical Mode Decomposition

Abstract

Electrical impedance tomography (EIT) imaging is a new kind of imaging technology that has developed in recent years. This technology has many advantages such as noninvasive, small volume and so on, but there are also many shortcomings such as imaging quality is not high, in order to improve the image quality, image processing methods is used in this paper. Significant different with the traditional method, empirical mode decomposition method does not require any prior basis function. It is an adaptive time-frequency analysis method. The two-dimensional empirical mode decomposition algorithm is adopted to process EIT images in this paper, and the compactly supported radial basis function interpolation algorithm is used for the post-processing of three organic glass rod EIT imaging. The image is decomposed into six IMFs in the experiment, some IMFs contain useful information, and the rest of them contains a lot of noise. The experimental results show that this method can significantly improve the EIT image quality.