Gait Recognition Using Pose Estimation and Signal Processing

8Citations
Citations of this article
12Readers
Mendeley users who have this article in their library.

This article is free to access.

Abstract

Gait is a biometry that differentiates individuals by the way they walk. Research on this topic has gained evidence since it is unobtrusive and can be collected at distance, which is desirable in surveillance scenarios. Most of the previous works have focused on human silhouette as representation. However, they suffer from many factors such as movement on scene, clothing and carrying conditions. To avoid such problems, this work employs pose estimation to retrieve the coordinates of body parts, which are transformed into signals and movement histograms to be used as feature descriptors. While the former descriptors are used with the Subsequence Dynamic Time Warping that compares signals from probe and gallery, the Euclidean distance is used on the latter to find the person on gallery that is closest to probe. Finally, the outputs of both are fused. This work was evaluated on all views of CASIA Dataset A and compared to existing ones, demonstrating its efficacy.

Cite

CITATION STYLE

APA

de Lima, V. C., & Schwartz, W. R. (2019). Gait Recognition Using Pose Estimation and Signal Processing. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11896 LNCS, pp. 719–728). Springer. https://doi.org/10.1007/978-3-030-33904-3_68

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free