Classification of Rigid Objects from High-Bandwidth Data Acquisition

Research in haptics has advanced for additional technologies such as remote operation and VR. Within this field, there’s a growing need for preserving environmental fidelity to achieve higher precision in tactile reproduction. This paper focuses on tactile feedback and discusses object classification inspired by methods for preserving hardness. Identifying the hardness of objects typically requires grasping, which necessitates a certain level of softness; hence, rigid objects cannot be identified. Moreover, when focusing solely on classification, some methods involve chemical techniques or sound-based tests like hammering. However, these methods are not designed for tactile feedback and do not directly correlate with tactile information. Therefore, this paper proposes a method for classifying previously untargeted rigid objects using a scheme similar to that for identifying object hardness. To accomplish this, experiments were conducted using a high-bandwidth data acquisition method with Field Programmable Gate Array (FPGA). The experiment involved identifying objects with similar appearances but differing masses, such as aluminum and iron, and comparing the results.