Look-to-Touch: A Vision-Enhanced Proximity and Tactile Sensor for Distance and Geometry Perception in Robotic Manipulation

TL;DR

The paper tackles the limited environmental perception of traditional tactile sensors by introducing a vision-enhanced dual-modality palm (V-T Palm) that mechanically switches between proximity sensing and tactile sensing via a sliding window. The approach combines a nonlinear dynamic proximity mapping with a photometric-stereo-based tactile geometry reconstruction, enabled by a belt-driven mode-switching mechanism and integrated with soft robotic fingers. Key contributions include full-scale distance sensing from $50 \mathrm{cm}$ to $-3 \mathrm{mm}$, nanometer-scale roughness detection, sub-millimeter 3D texture reconstruction, and demonstrated improvements in grasping efficiency and in-hand manipulation. The work has practical impact for robust, compact perception in unstructured environments, enabling proactive grasp planning, texture-based object differentiation, and fine intra-hand adjustments.

Abstract

Camera-based tactile sensors provide robots with a high-performance tactile sensing approach for environment perception and dexterous manipulation. However, achieving comprehensive environmental perception still requires cooperation with additional sensors, which makes the system bulky and limits its adaptability to unstructured environments. In this work, we present a vision-enhanced camera-based dual-modality sensor, which realizes full-scale distance sensing from 50 cm to -3 mm while simultaneously keeping ultra-high-resolution texture sensing and reconstruction capabilities. Unlike conventional designs with fixed opaque gel layers, our sensor features a partially transparent sliding window, enabling mechanical switching between tactile and visual modes. For each sensing mode, a dynamic distance sensing model and a contact geometry reconstruction model are proposed. Through integration with soft robotic fingers, we systematically evaluate the performance of each mode, as well as in their synergistic operation. Experimental results show robust distance tracking across various speeds, nanometer-scale roughness detection, and sub-millimeter 3D texture reconstruction. The combination of both modalities improves the robot's efficiency in executing grasping tasks. Furthermore, the embedded mechanical transmission in the sensor allows for fine-grained intra-hand adjustments and precise manipulation, unlocking new capabilities for soft robotic hands.

Figures (15)

• Figure 1: Illustration of two working modes of the V-T Palm and its application in a robotic system.
• Figure 2: Design of V-T palm. (a) Transmission mechanism is introduced to integrate the sensing functions of two modes into a compact structure. (b) Detailed composition structure.
• Figure 3: Measurement principle. The proximity sensing mode utilizes the camera to continuously capture images of external target objects. Under tactile mode, the LED strip is activated, enabling the acquisition of tactile information through the internal light field.
• Figure 4: Comparison of different illumination methods. (a) and (b) show the comparison of light intensity distribution between the direct illumination scheme and the diffuse reflection scheme of lamp beads simulated in Blender. (c) illustrates the actual illumination condition of the sensor.
• Figure 5: Experimental platform setup of proximity sensing. (a) illustrates the detailed composition of the test platform; (b) and (c) show two kinds of disc steering gear couplings with inner diameters of 5 mm and 35 mm, which realize the movement of the target object at different speeds.