A cryogenic visible calibration and image evaluation facility (VCCIEF) was constructed to assess the effectiveness of electrical capacitance tomography systems in cryogenic conditions

known as Cryo-ECT. This facility was utilized to conduct dynamic

real-time imaging trials with liquid nitrogen (LN2). The actual flow patterns were captured using a camera and contrasted with the imaging outcomes. The capacitance data collected from these experiments were subsequently processed using three distinct methods: linear back projection

Landweber iteration

a fully connected deep neural network

and a convolutional neural network. This allowed for a comparative analysis of the performance of these algorithms in practical scenarios. The findings from the LN2 experiments demonstrated that the Cryo-ECT system

when integrated with the VCCIEF

was capable of successfully executing calibration

generating flow patterns

and performing imaging tasks. The system provided observable

clear

and precise phase distributions of the liquid nitrogen-vaporous nitrogenflow within the pipeline.