Estimation de mesures anthropométrique avec une seule image

Abstract

In this research we propose a method to obtain anthropometric measurements of a subject from a single image by scaling a morphological and normalized avatar of the subject to the real world. To acquire the scaling factor, we leverage the distance at which the subject is placed from the camera. We therefore propose a method to estimate the distance between the camera and the subject from a monocular image without restricting the orientation of the head. The inter pupillary distance of the subject is used as an anchor to estimate the size of pixels in the plane of the subject and the distance to the camera is recovered from this measurement. The inter pupillary distance in the image is estimated by using a learning-based approach that detects the head of the subject and estimates the orientation of the head with respect to the optical axis of the camera. It also proposes a method to estimate the pupillary distance from two images in case this information is not available. The proposed approaches are validated by comparing the results obtained with an OptiTrack motion-capture system. It is demonstrated that using the proposed algorithm allows a more robust estimation of the distance regardless of the head rotation. It also shows that the estimation of the pupillary distance is possible using only two images. Lastly, the research provides successful results on providing a scaled avatar that can be used for anthropometric measurements.

Date	31 Aug 2022
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Carlos Vázquez (Supervisor)