Traditional electrical tomography tactile sensors consider the usage of the system’s finite element model. This approach brings disadvantages that jeopardise their applicability aspect and wide use. To address this limitation, the main thrust of this work is to present a method for touch position identification for an electrical tomography flexible tactile sensor. This is done by using a supervised machine learning algorithm for performing classification, namely quadratic discriminant analysis. This approach provides accurate contact location identification, increasing the detection speed and the sensor versatility when compared to traditional electrical tomography approaches. Results obtained show classification accuracy rates of up to 91.6% on unseen test data and an average euclidean error ranging from 1 to 10 mm depending on the contact location over the sensor. The sensor is then applied in real case scenarios to show its efficiency. These outcomes are encouraging since they promote the future practical usage of flexible and low-cost sensors.
Touch Position Detection in Electrical Tomography Tactile Sensors Through Quadratic Classifier
Carbonaro, Nicola;Tognetti, Alessandro
2019-01-01
Abstract
Traditional electrical tomography tactile sensors consider the usage of the system’s finite element model. This approach brings disadvantages that jeopardise their applicability aspect and wide use. To address this limitation, the main thrust of this work is to present a method for touch position identification for an electrical tomography flexible tactile sensor. This is done by using a supervised machine learning algorithm for performing classification, namely quadratic discriminant analysis. This approach provides accurate contact location identification, increasing the detection speed and the sensor versatility when compared to traditional electrical tomography approaches. Results obtained show classification accuracy rates of up to 91.6% on unseen test data and an average euclidean error ranging from 1 to 10 mm depending on the contact location over the sensor. The sensor is then applied in real case scenarios to show its efficiency. These outcomes are encouraging since they promote the future practical usage of flexible and low-cost sensors.File | Dimensione | Formato | |
---|---|---|---|
EIT_Classification_with_ML___IEEE_Sensor_Journal___v2__Copy_.pdf
accesso aperto
Tipologia:
Documento in Pre-print
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
21.73 MB
Formato
Adobe PDF
|
21.73 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.