Wyniki wyszukiwania - BazTech

1

System analizy obrazu dla potrzeb systemu AR

Garbat Piotr, Galiński Grzegorz, Bieniek Mateusz, Żakowski Michał, Lasocki Maciej

Przegląd Telekomunikacyjny + Wiadomości Telekomunikacyjne

|

2022

|

nr 4

143--146

PL

Przedstawiono ideę systemu wizyjnego umożliwiającego analizę obrazów z wielu kamer dla potrzeb systemów rzeczywistości wspomaganej. Przedstawione rozwiązanie zapewnia efektywne wyznaczanie pozycji i ścieżek ruchu poruszających się osób w przestrzeni miejskiej oraz ich aktywności. W artykule przedstawiono kluczowe moduły analizy obrazu z wielu widoków, w tym: podsystem wykrywania i śledzenia pieszych (MOT) oraz podsystem wykrywania aktywności użytkowników. Omawiane podsystemy zostały przebadane w celu określenia efektywności zaproponowanych algorytmów określania ścieżki ruchu i rozpoznawania aktywności.

EN

This article presents the idea of a vision system that allows for the analysis of images from multiple cameras for the purposes of augmented reality systems. The proposed solution effectively determines the position and paths of movement of moving people in the urban space and their activities. The article presents the critical modules of the image analysis from multiple views, including the pedestrian detection and tracking (MOT) subsystem and the user activity detection subsystem. The subsystems were stud to determine the effectiveness of the proposed algorithms for determining the traffic path and recognizing activity.

2

Action recognition based on shape features and their correlation

Gościewska K.

Journal of Theoretical and Applied Computer Science

|

2016

|

Vol. 10, nr 1

11--18

EN

This paper presents an approach for action recognition based on binary silhouette sequences extracted from consecutive frames of a video. It uses shape descriptors and correlation coefficient to represent and match entire sequences, regardless the number of frames. Each set of binary silhouettes corresponds to one action, such as jumping or waving. The paper provides experimental results on the use of the proposed approach and four shape description algorithms, namely the Two-Dimensional Fourier Descriptor, Generic Fourier Descriptor, Point Distance Histogram and UNL-Fourier Descriptor. The results are analysed in terms of the highest classification accuracy and the smallest shape descriptor size.

3

An RDF-based action recognition framework with feature selection capability, considering therapy exercises utilizing depth cameras

Negin F., Akgül C.B., Yüksel K.A., Erçil A.

Journal of Theoretical and Applied Computer Science

|

2014

|

Vol. 8, nr 3

3--22

EN

Emerging cost-efficient depth sensor technologies reveal new possibilities to cope with difficulties in action recognition. Depth information improves the quality of skeleton detection process, hence, pose estimation can be done more efficiently. Recently many studies focus on temporal analyses over estimated skeleton poses to recognize actions. In this paper we have an inclusive study of the spatiotemporal kinematic features and propose an action recognition framework with feature selection capability to deal with the multitudinous of features by leveraging data mining capabilities of random decision forests. We describe human motion via a rich collection of kinematic feature time-series computed from the skeletal representation of the body in motion. We discriminatively optimize a random decision forest model over this collection to identify the most effective subset of features, localized both in time and space. Later, we train a support vector machine classifier on the selected features. This approach improves upon the baseline performance obtained using the whole feature set with a significantly less number of features (one tenth of the original). To justify our method we test the framework on various datasets and compared it with state-of-theart. On MSRC-12 dataset [25] (12 classes), our method achieves 94% accuracy. On the WorkoutSU-10 dataset [28], collected by our group (10 physical exercise classes), the accuracy is 98%. On MSR Action3D dataset [9] (20 classes) we obtain 87% average accuracy and for UTKinect-Action dataset [10] (10 classes) the accuracy is 92%. Other than regular activities, we also tried our approach to detect a falling person using the dataset which we recorded as an extension to our original dataset. We test how our method adjusts on different types of actions and we obtained promising results for this type of action. We discuss that our approach provides insights on the spatiotemporal dynamics of human actions and can be used to as part of different applications especially for rehabilitation of patients.

4

Extraction of Discriminative Patterns from Skeleton Sequences for Accurate Action Recognition

Thanh T. T., Chen F, Kotani K., Le B.

Fundamenta Informaticae

|

2014

|

Vol. 130, nr 2

247--261

EN

Emergence of novel techniques devices e.g., MS Kinect, enables reliable extraction of human skeletons from action videos. Taking skeleton data as inputs, we propose an approach to extract the discriminative patterns for efficient human action recognition. Each action is considered to consist of a sequence of unit actions, each of which is represented by a pattern. Given a skeleton sequence, we first automatically extract the key-frames, and then categorize them into different patterns. We further use a statistical metric to evaluate the discriminative capability of patterns, and define them as local features for action recognition. Experimental results show that the extracted local descriptors could provide very high accuracy in the action recognition, which demonstrate the efficiency of our method in extracting discriminative unit actions.

5

A neural network approach to recognition of the selected human motion patterns

Mikrut Z., Smoleń M.

Automatyka / Akademia Górniczo-Hutnicza im. Stanisława Staszica w Krakowie

|

2011

|

T. 15, z. 3

535-543

EN

In this paper we proposed a representation of the activities performed by humans in the form of a histogram of optical flow directions. Histogram was calculated in the mask of object contour and aggregated to the eight bins. The data set was based on analysis of the video clip, in which four people were performing nine activities, such as walking, sitting and rising from a chair or reaching (up and forward). To recognize the performed activity the backpropagation neural network with one hidden layer was used. The recognition results varied from 80 to 88% for individuals. It was found that it is not possible to identify a person's activities using the network trained by data of another person.

PL

W artykule zaproponowano metodę reprezentacji czynności wykonywanej przez człowieka w postaci histogramu kierunków pola ruchu. Histogram był obliczany w masce konturu sylwetki i agregowany do ośmiu kierunków. Zbiór danych powstał na podstawie analizy filmu, na którym cztery osoby wykonywały 9 czynności, takich jak chodzenie, siadanie i wstawanie z krzesła czy też sięganie (w górę i do przodu). Do rozpoznawania wykorzystano sieć neuronową typu backpropagation z jedną warstwą ukrytą. Osiągnięto wyniki rozpoznawania na poziomie 80-88% dla pojedynczych osób. Stwierdzono, że nie jest możliwe rozpoznawanie czynności danej osoby za pomocą sieci nauczonej danymi innej osoby.

6

Recognition of actions in meeting videos using timed automata

Pelc L., Kwolek B.

Machine Graphics and Vision

|

2006

|

Vol. 15, No. 3/4

577-584

EN

This paper addresses the problem of action recognition in meeting videos. A declarative knowledge provided graphically by the user together with person positions extracted by a tracking algorithm are used to generate the data for recognition. The actions have been formally specified using timed automata. The specification was verified on the basis of simulation tests as well as an analysis. The tracking is accomplished using a particle filter built on cues such as color, gradient and shape.