Capture and Modelling of the Shod Foot in Motion – CaMoPi

The proposed project aims at combining in-depth and surface motion capture using a hybrid video / X-ray platform to produce a new type of data. An estimate of the position of the skeletal structure will be computed from X-ray imaging and linked to the surface mesh obtained by video. This 4D model will provide shoe manufacturers with a first glimpse of skeletal movement over time.

From this data, the multi-body modeling coupled to a finite element method will give key information on the state of the foot. The complexity of the anatomical structure of the foot being high in terms of number of elements (26 bone segments, 33 joints, 107 ligaments, 20 muscles), a simplified substitution model will also be developed for practical industrial exploitation.

This will focus on improving «lasts« (the models around which shoe prototypes are built), a critical step in shoe design. The project aims to streamline the process by providing metrics and models that will speed up the process and potentially reduce the number of trial and error prototyping steps. Two key aspects of shoe design will be optimized: the biomechanics of movement and the comfort of shoes.
Overall, the goal is to help manufacturers validate their lasts in an effective and scientific way in relation to the desired objectives (performance, comfort, ...) in order to accelerate the design of shoes and to reduce the costs.

No major results yet to be publicized.

The project is expected to produce unprecedented results showing how the internal movement of the foot is structured. In fact, beyond the specific industrial perspectives in shoe design, the proposed capture and modeling methods should have a variety of benefits.
For example, many applications are expected in the medical field, in the treatment of osteoarthritis of the knee (orthopedic insoles), osteoporosis (dynamic movement therapy) or the prevention of deformation of the foot bone (e.g. hallux valgus and quintus varus). Many applications are also possible in sports science.

1. Julien Pansiot and Edmond Boyer. CT from Motion: Volumetric Capture of Moving Shapes with X-rays and Videos. In British Machine Vision Conference (BMVC), London, September 2017. hal.inria.fr/hal-01585344
This publication proposes a new method of dense 3D reconstruction of a free-motion sample using an X-ray imager and multiple color cameras. This method is the first technical brick of the CaMoPi project.

2. Julien Pansiot and Edmond Boyer. CBCT of a Moving Sample from X-rays and Multiple Videos. IEEE Transactions on Medical Imaging (TMI). 2018. hal.inria.fr/hal-01857487

The main objective of the CaMoPi project is to capture and model dynamic aspects of the human foot with and without shoes. To this purpose, video and X-ray imagery will be combined to generate novel types of data from which major breakthroughs in foot motion modelling are expected.
Given the complexity of the internal foot structure, little is known about the exact motion of its inner structure and the relationship with the shoe. Hence the current state-of-the art shoe conception process still relies largely on ad-hoc know-how. This project aims at better understanding the inner mechanisms of the shod foot in motion in order to rationalise and therefore speed up and improve shoe design in terms of comfort, performance, and cost. This requires the development of capture technologies that do not yet exist in order to provide full dense models of the foot in motion. To this aim, the main directions of work within CaMoPi will be articulated around the following three main axes.
1. Combined surface and volumetric motion capture using a hybrid video / X-ray platform to produce new data type on moving shapes. Novel methods will be devised for fusing both imaging modalities within a single coherent 4D model encompassing surface and in-depth volumetric data.
2. Accurate modelling of the foot in motion. A physical in-depth comprehension of the foot motion using multi-body modelling coupled with a finite element method (FEM) will give key information on the foot state. A simplified surrogate model will also be developed.
3. Industrial validation with shoe manufacturers. The new data produced will be analysed by industrials. Key features extracted from the surrogate model will be integrated in the shoe design process. Prototype models will be tested again within the capture system to quantify the improvement, closing effectively the feed-back loop between research and industrial design.
The proposed project implements a full pipeline from data capture and modelling to practical industrial exploitation with CTC, which performs R&D for a broad pool of industrial affiliates. Hence we expect the outcomes in terms of novel data types, capture procedures, and modelling to impact the shoe manufacturing process and more broadly lower limb biomechanics.