Trackability evaluation of reconstruction algorithms to the change of measured objects in electrical tomography

The continuous monitoring of the changing process is an important application field of electrical tomography (ET). In this changing process, the size, position and shape of measured objects are always alterative. The trackability of algorithms to the change of measured objects is important to the application of ET. The single object model group and two-object-model group were established to simulate the change of measured objects. The single object model group includes the circle model group and square model group. The suitable evaluation parameters were designed to evaluate the trackability of the different algorithms quantitatively, which includes the single image parameter and group parameter. Evaluation software was developed, which can generate measured boundary data of different models, complete reconstructed image greying, calculate evaluation parameters and plot parameter curves, etc. Furthermore, the trackability of ten selected algorithms was evaluated by this evaluation software. The results show that the trackability of the different algorithms is different in the evaluation of the different model group. Therefore, the different model group should be established according to the application requirement. Then the suitable algorithm for a particular application could be chosen through the evaluation process.