The use and development of plantar pressure measurement systems (PPMs) have increased in recent years, in order to understand and objectively evaluate the interaction between the feet and the support surfaces. The PPMs are mainly made with resistive sensor arrays, which have problems associated with crosstalk reading errors. Several circuits aimed at reducing the crosstalk effect have been investigated; however, none of these circuits were evaluated in large-area matrices or practical applications. Thus, in this work, we explain the development and characterization of the LiebScan system, which is a stationary PPMs with 2304 sensels based on the improved isolated drive feedback circuit for crosstalk control. The commercial equipment recommended for PPM linearization and characterization of a large-area matrix is a high-cost pneumatic device. In this sense, we propose a characterization by areas using a universal testing machine and an artificial neural network, since each sensel has an individual and unique response. Although the LiebScan system supports pressures up to 300 kPa, it presents competitive characteristics when compared with commercial equipment, in terms of accuracy, hysteresis, creep, crosstalk, and center of pressure processing. The most remarkable feature among these cited before is the low crosstalk effect, with a mean error of 0.23%. The pressures measured because of this crosstalk error are below the pressure threshold, thus, the crosstalk effect is practically null. This work describes the design, development, characterization and detailed evaluation of a PPMs with commercial features.
Read full abstract