Dynamic 3D shape of the plantar surface of the foot using coded structured light: a technical report

Ali Thabet, Emanuele Trucco, Joaquím Salví, Weijie Wang, Rami AbboudJournal of Foot and Ankle Research 2014 · DOI

Abstract

The results obtained in the experiments show positive accuracy and repeatability results when compared to current literature. The design also shows to be superior to the systems available in the literature in several factors. Further studies need to be done to quantify the reliability of the system in clinical environments.

Introduction

Traditional foot scanning is typically a static process. However, a foot at rest differs significantly from a foot in motion, which is critical for designing effective custom orthotics and footwear. Existing dynamic 3D reconstruction systems often suffer from high costs (ranging from $5,000 to $50,000 USD) or require intrusive markers, special socks, or skin paint that can alter a subject's natural gait.

To address these issues, researchers developed the 4D Foot Reconstruction System (4DFRS). This system captures the moving plantar surface of a foot during a single step with sub-millimeter accuracy, providing a completely natural "walk-over" experience without external markers.

Methodology

The 4DFRS utilizes a hardware setup elegantly integrated into a walkway. A single high-definition camera and a 3-LCD projector are housed in a pit beneath a transparent, toughened glass surface.

The hardware setup of the 4DFRS, featuring a camera and projector recessed beneath the walkway.

The core technology is Coded Structured Light (CSL). Instead of relying on natural skin texture, the system projects a specific pattern of vertical colored stripes onto the sole of the foot.

  1. Acquisition: The camera records the distortion of these stripes as the foot moves across the glass.
  2. Processing: Depth is calculated based on the known geometry of the projected stripes.
  3. Reconstruction: The software generates a dense 3D point cloud for every video frame at 60 Hz, creating a dynamic 3D model of the foot's contact and roll-off.
CSL projector data showing the light pattern (left) and its application to a foot surface (right).

To compare scans across different sessions, the Iterative Closest Point (ICP) algorithm was employed to align 3D shapes and measure deviations.

Experimental Design

The system was evaluated using two primary datasets:

  • Engineering Test Objects: A rigid wooden plane (30 cm x 15 cm) was imaged at various orientations to establish baseline accuracy.
  • Clinical Subject Group: 27 healthy subjects participated in the study. Researchers captured four static and four dynamic sequences for both feet, with a follow-up session one week later to assess repeatability.

Key Results

The 4DFRS demonstrated high precision and consistency:

  • Static Accuracy: Mean error of 0.3 mm when imaging a flat plane.
  • Clinical Repeatability: Average differences between weekly trials were 2.4 mm for static scans and 2.8 mm for dynamic scans.
  • Temporal Resolution: The system successfully captured the full stance phase, providing 20 to 40 frames of high-density 3D data per step.

Discussion and Limitations

The 4DFRS offers a non-invasive, computationally efficient solution for dynamic foot analysis. By recessing the technology, it minimizes the "Hawthorne effect," ensuring more natural movement data.

However, several limitations remain:

  • Field of View: The system only captures the plantar (bottom) surface, not the dorsal surface or ankle.
  • Optical Distortion: Higher errors occurred at the image periphery due to standard lens distortion.
  • Workflow: Dynamic trials currently require manual frame selection for synchronization.
  • Clinical Validation: Future work is needed to validate the system's performance in patients with severe deformities or pathological gaits.

Summary updated

Abstract

The results obtained in the experiments show positive accuracy and repeatability results when compared to current literature. The design also shows to be superior to the systems available in the literature in several factors. Further studies need to be done to quantify the reliability of the system in clinical environments.
BibTeX
@article{thabet2014dynamic,
  author = {Ali Thabet and Emanuele Trucco and Joaquím Salví and Weijie Wang and Rami Abboud},
  title = {{Dynamic 3D shape of the plantar surface of the foot using coded structured light: a technical report}},
  journal = {Journal of Foot and Ankle Research},
  year = {2014},
  doi = {10.1186/1757-1146-7-5},
}
Figure