Algorithm For Determining A Person’S Posture In A Video Sequence

R.K. Uskenbayeva, А.А. Куандыков, Г.Г. Аубакирова, G.K. Kuatbayeva, B. T. Kossynbayev, A. A. Kuatbayeva, B.M. Aubakirova

2020

Closed access Metadata (CSL-JSON)
DOI 10.1109/cbi49978.2020.10063
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

This article discusses the problems of recognition and determination of the pose of one person, several on the image and video sequence. To solve this problem, the characteristic features of human image recognition algorithms are analyzed- as a set of points and lines connecting them. We can imagine a person as a set of dynamic fixtures and parts connecting them. In the most elementary models, they are radius vectors between two points. Two publicly available algorithms for recognizing a person's pose AlphaPose and OpenPose- were also considered [16; 17]. Based on the results of the study, each algorithm had advantages in which areas.