Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Embedded Object Detection and Mapping in Soft Materials Using Optical Tactile Sensing

Jose A. Solano-Castellanos, Won Kyung Do, Monroe Kennedy

TL;DR

The paper addresses locating and reconstructing embedded rigid objects beneath a soft, opaque medium using an optical tactile sensor. It introduces a two-stage framework: an exploration phase that builds a probabilistic map of object locations via Bayesian Gaussian Process classification, and a mapping phase that concentrates samples in high-probability regions to recover the underlying topography. The approach is validated on an experimental setup with quartz beads under a polyethylene foam, showing the method outperforms random sampling in both phases and achieving consistent, accurate shape estimation. This work enables non-visual tactile inspection of heterogeneous soft materials with potential applications in packaging, medical palpation, and haptic prosthetics, and lays groundwork for incorporating additional cues such as stress and 3D/curved surfaces in future work.

Abstract

In this paper, we present a methodology that uses an optical tactile sensor for efficient tactile exploration of embedded objects within soft materials. The methodology consists of an exploration phase, where a probabilistic estimate of the location of the embedded objects is built using a Bayesian approach. The exploration phase is then followed by a mapping phase which exploits the probabilistic map to reconstruct the underlying topography of the workspace by sampling in more detail regions where there is expected to be embedded objects. To demonstrate the effectiveness of the method, we tested our approach on an experimental setup that consists of a series of quartz beads located underneath a polyethylene foam that prevents direct observation of the configuration and requires the use of tactile exploration to recover the location of the beads. We show the performance of our methodology using ten different configurations of the beads where the proposed approach is able to approximate the underlying configuration. We benchmark our results against a random sampling policy.

Embedded Object Detection and Mapping in Soft Materials Using Optical Tactile Sensing

TL;DR

The paper addresses locating and reconstructing embedded rigid objects beneath a soft, opaque medium using an optical tactile sensor. It introduces a two-stage framework: an exploration phase that builds a probabilistic map of object locations via Bayesian Gaussian Process classification, and a mapping phase that concentrates samples in high-probability regions to recover the underlying topography. The approach is validated on an experimental setup with quartz beads under a polyethylene foam, showing the method outperforms random sampling in both phases and achieving consistent, accurate shape estimation. This work enables non-visual tactile inspection of heterogeneous soft materials with potential applications in packaging, medical palpation, and haptic prosthetics, and lays groundwork for incorporating additional cues such as stress and 3D/curved surfaces in future work.

Abstract

In this paper, we present a methodology that uses an optical tactile sensor for efficient tactile exploration of embedded objects within soft materials. The methodology consists of an exploration phase, where a probabilistic estimate of the location of the embedded objects is built using a Bayesian approach. The exploration phase is then followed by a mapping phase which exploits the probabilistic map to reconstruct the underlying topography of the workspace by sampling in more detail regions where there is expected to be embedded objects. To demonstrate the effectiveness of the method, we tested our approach on an experimental setup that consists of a series of quartz beads located underneath a polyethylene foam that prevents direct observation of the configuration and requires the use of tactile exploration to recover the location of the beads. We show the performance of our methodology using ten different configurations of the beads where the proposed approach is able to approximate the underlying configuration. We benchmark our results against a random sampling policy.
Paper Structure (9 sections, 14 equations, 9 figures, 1 algorithm)

This paper contains 9 sections, 14 equations, 9 figures, 1 algorithm.

Table of Contents

  1. Introduction
  2. Extracting Local Information Through Touch
  3. Active Exploration for Shape Reconstruction
  4. Exploration and Mapping
  5. Exploration
  6. Mapping
  7. Experimental Setup
  8. Results and Discussion
  9. Conclusion

Figures (9)

  • Figure 1: Embedded Object Detection. In this work we demonstrate the ability to detect and map objects embedded in a soft medium. This work has future applications to medical automated assistance for patient palpation.
  • Figure 2: Graphical representation of the proposed method. In the exploration phase (E.1 through E.6) the probable areas of hard embedded objects below the soft surface are estimated. In the mapping phase (M.1 and M.2) a more thorough interaction of such areas is conducted to approximate the underlying topology.
  • Figure 3: Relation of the proposed method with the Sense-Plan-Act cycle of robotics.
  • Figure 4: Experimental configuration for the detection and mapping of embedded rigid objects in soft material matrices. (a) A perforated acrylic sheet allows to accommodate quartz bead clusters in any desired configuration. (b) The beads are covered with a layer of polyethylene foam and secured in place with two longitudinal acrylic strips. (c) Side by side of the raw image from camera and the resulting depth image from DenseTact 2.0.
  • Figure 5: Configurations of the quartz beads below the foam for the evaluation of the performance of the proposed method
  • ...and 4 more figures