Method for calibrating industrial CT system detector by group series winding

Abstract

Calibration of industrial CT system inspector by group series winding relates to re-established images. It carries out to process projection data generated by inspector group series winding model based on Ordered Subsets Convex, to make medium filtration after each iteration in order to speed up convergence. This invention needs less times of convergence and has quality calibration.

Description

(1) Technical field [0001] The invention belongs to the use of mathematical methods to correct the reconstructed image of the CT system in the nuclear technology field. (2) Background technology [0002] In the actual industrial CT system, every 8 detectors form a group, and a group of detectors cross each other, so there are obvious artifacts in the reconstructed image. At present, there are three main methods for correcting the effect of detector cross-winding in CT systems: using direct deconvolution functions, deconvolution iterative methods, and adding a detector cross-winding model to the iterative reconstruction. The above methods are not aimed at the case of packet crosstalk. They have the following shortcomings when they are used in our actual CT system detector packet crosstalk: [0003] 1) For the correction method using the direct deconvolution function, because the deconvolution function always has errors, the correction of each set of data is inaccurate, so the imag...

AI Technical Summary

Problems solved by technology

[0003] 1) For the method of correcting with the direct deconvolution function, because the deconvolution function always has errors, the correction of each set of data is not accurate, so the image reconstructed by recombining each set of data still has obvious artifacts. Shadow, poor correction effect

[0004] 2) For the deconvolution iterative method, it is not practical for industrial CT systems with a large amount of data because it needs to be iterated, and generally it takes multiple iterations to achieve the purpose of correction

[0006] 4) For the method of adding the detector cross-winding model in iterative reconstruction, it is currently limited to the correction of general detector cross-winding, and the correction speed is slow, requiring multiple iterations to complete

Moreover, many of the current iterative methods converge very slowly, which is not applicable to the actual system.

Images