TIRgel: A Visuo-Tactile Sensor With Total Internal Reflection Mechanism for External Observation and Contact Detection

Shixin Zhang,Yuhao Sun,Jianhua Shan,Zixi Chen,Fuchun Sun,Yiyong Yang,Bin Fang
DOI: https://doi.org/10.1109/lra.2023.3306670
IF: 5.2
2023-10-01
IEEE Robotics and Automation Letters
Abstract:This letter proposes a vision-based tactile sensor named TIRgel, leveraging visual integration to simplify the sensing system. First, the sensor achieves conversion of visual and tactile modality via focus adjustment. Under far-focus imaging, the camera can observe external environments; Under near-focus imaging, we perform total internal reflection (TIR) within the lenticular elastomer to visualize tactile features. The effectiveness of the sensing mechanism is verified via MPM (Material Point Method) simulation. Second, the TIRgel can capture and extract micron-scale tactile details based on luminosity information. Tactile imaging has the natural fusion of visual and tactile information, where the luminosity boundary reflects the texture feature, and the background contains the object's appearance information. Inherent data fusion simplifies multimodal data acquisition and reduces the information processing difficulty. Finally, the TIRgel can resist the interference of ambient brightness via internal illumination and has a higher adaptive capacity to the environment than unidirectional perspective VBTSs. We believe that it has the potential to provide visual-tactile integration for robot operation.
robotics
What problem does this paper attempt to address?