6533b85cfe1ef96bd12bd23d

RESEARCH PRODUCT

Quantifying Mean Shape and Variability of Footprints Using Mean Sets

Enrique AlcántaraE. De VesGuillermo AyalaJ. DomingoBeatriz NacherE. DiazÁLvaro Page

subject

Set (abstract data type)FootprintComputer scienceRobustness (computer science)business.industryPattern recognitionNoise (video)RepeatabilityArtificial intelligenceRaw databusinessHighly sensitive

description

This paper1 presents an application of several definitions of a mean set for use in footwear design. For a given size, footprint pressure images corresponding to different individuals constitute our raw data. Appropriate footwear design needs to have knowledge of some kind of typical footprint. Former methods based on contour relevant points are highly sensitive to contour noise; moreover, they lack repeatability because of the need for the intervention of human designers. The method proposed in this paper is based on using mean sets on the thresholded images of the pressure footprints. Three definitions are used, two of them from Vorob’ev and Baddeley-Molchanov and one morphological mean proposed by the authors. Results show that the use of mean sets improves previous methodologies in terms of robustness and repeatability.

https://doi.org/10.1007/1-4020-3443-1_41