Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: karate

Sortuj według:

Ogranicz wyniki do:

The application of topological data analysis to human motion recognition

Żelawski Marcin, Hachaj Tomasz

Technical Transactions

2021

Vol. 118, iss. 1

art. no. e2021011

Human motion analysis is a very important research topic in the field of computer vision, as evidenced by a wide range of applications such as video surveillance, medical assistance and virtual reality. Human motion analysis concerns the detection, tracking and recognition of human activities and behaviours. The development of low-cost range sensors enables the precise 3D tracking of body position. The aim of this paper is to present and evaluate a novel method based on topological data analysis (TDA) for motion capture (kinematic) processing and human action recognition. In contrast to existing methods of this type, we characterise human actions in terms of topological features. The recognition process is based on topological persistence which is stable to perturbations. The advantages of TDA are noise resistance and the ability to extract global structure from local information. The method we proposed in this paper deals very effectively with the task of human action recognition, even on the difficult classes of motion found in karate techniques. In order to evaluate our solution, we have performed three-fold cross-validation on a data set containing 360 recordings across twelve motion classes. The classification process does not require the use of machine learning and dynamical systems theory. The proposed classifier achieves a total recognition rate of 0.975 and outperforms the state-of-theart methods (Hachaj, 2019) that use support vector machines and principal component analysis-based feature generation.

Computer – aided method for lower limbs kinematic analysis

Ogiela Marek R., Hachaj Tomasz

Przegląd Elektrotechniczny

2020

R. 96, nr 6

15--25

The aim of this paper is to propose a novel method that enables kinematic analysis of motion capture (MoCap) data of lower limbs activities by comparison of body joints trajectories to the reference template. We propose an appropriate human body kinematic model, MoCap aligning procedure and heuristic evaluation with Dynamic Time Warping (DTW) - based approach. In contrast to other state-of-the-art papers, where analysis is performed on the single joint on the selected two-dimensional plane, we performed three-dimensional evaluation of human body by analyzing the whole kinematic chain jointly. This approach allows us to find which body joints affected the difference between the input and reference recordings the most. This is valuable information that a person who evaluates MoCap data expects to find. We have also performed kinematic analysis applying commonly used kinematic parameters proposed in state-of-the-art researches in order to show that in our case, when there is no restriction on speed or dynamic of action to be analyzed, those parameters cannot be used to draw valuable conclusions. We have tested our method on a dataset consisting recordings of four karate athletes with various experience in Shorin Ryu karate school. While comparing our algorithm’s results to experts evaluation the true positive rate equals 0.93 while negative rate 0.96.

W pracy zaproponowano nową metodę analizy kinematyki kończyn dolnych przy pomocy nagrań motion capture (MoCap). Zaproponowane rozwiązanie pozwala na kompleksową analize całosci łańcucha kinematycznego. Przetestowaliśmy zaproponowany algorytm na zbiorze danych zawierającym nagrania czterech zawodników Shorin Ryu karate uzyskując zadawalające wyniki w porównaniu do analogicznej ewaluacji przeprowadzonej przez eksperta.