| Abstract: | A new sparse tomography method for observing the rapid internal microstructure evolution of material, called the Algebraic Filtered-Back-Projection and Total Variation Minimization (AFBP-TVM) iteration sparse reconstruction algorithm, was proposed in this paper. The new algorithm was developed by combining the two techniques of the Algebraic Reconstruction Technique (ART) and the Filtered-Back-Projection (FBP) on the basis of analysis in linear space. A series of numerical reconstruction experiments were conducted to validate the new algorithm. The results indicated the new algorithm can obtain satisfactory reconstruction images from 1/6 of the projections that were used in traditional algorithms. So the time spent on projection acquisition process can be reduced to 1/6 of that in traditional tomography method. The quality of images reconstructed by new algorithm was better than other algorithms, which was evaluated by three quantitative parameters. The normalized average absolute distance criterion and the normalized mean square criterion, which were used to evaluate the relative error of the reconstruction results (smaller value means better quality of reconstruction), decreased from 0.3758 to 0.1272 and from 0.1832 to 0.0894 respectively. The standardized covariance criterion, which was used to evaluate the similarity level (greater value means higher accuracy of reconstruction), increased from 92.72% to 99.30%. Finally, the new algorithm was validated under actual experimental conditions. The results indicated that the AFBP-TVM algorithm obtained better reconstruction quality than other algorithms. It meant that the AFBP-TVM algorithm may be a suitable method for in situ investigation on material's rapid internal microstructure evolution in extreme complex environment. |
