Skip to Main Content
This paper introduces guided-path tomography (GPT) as a method for imaging on nonplanar surfaces by taking measurements at their periphery. The theory and practical implementation (hardware and software) of GPT is illustrated in the case of temperature mapping. The temperature distribution is obtained from dc measurements of the temperature-induced resistance changes (accuracy ±0.02%) in a grid of a small number of noninteracting transducers forming the GPT sensor. Images of the temperature distribution around a heated tip and across a thermally nonhomogenous flow are reconstructed using the additive algebraic reconstruction technique (AART). Results from applying several strategies for the design of the sensor are shown and discussed. Problems concerning the general applicability of GPT in its variants, as well as the improvement of its current implementation are discussed in detail and some industrial applications of GPT temperature mapping are suggested. It is shown that using a GPT temperature imaging sensor it is possible to generate Radon-transformed "phantom" data from electrical measurements.