Tomographic Proximity Imaging Using Conductive Sheet for Object Tracking.

This paper proposes a proximity imaging sensor based on a tomographic approach with a low-cost conductive sheet. Particularly, by defining capacitance density, physical proximity information is transformed into electric potential. A novel theoretical model is developed to solve the capacitance density problem using the tomographic approach. Additionally, a prototype is built and tested based on the model, and the system solves an inverse problem for imaging the capacitance density change that indicates the object's proximity change. In the evaluation test, the prototype reaches an error rate of 10.0-15.8% in horizontal localization at different heights. Finally, a hand-tracking demonstration is carried out, where a position difference of 33.8-46.7 mm between the proposed sensor and depth camera is achieved at 30 fps.