• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권2호
• /
• pp.929-944
• /
• 2019

Many critical applications require accurate real-time human action recognition. However, there are many hurdles associated with capturing and pre-processing image data, calculating features, and classification because they consume significant resources for both storage and computation. To circumvent these hurdles, this paper presents a recognition machine learning (ML) based system model which uses reduced data structure features by projecting real 3D skeleton modality on virtual 2D space. The MMU VAAC dataset is used to test the proposed ML model. The results show a high accuracy rate of 97.88% which is only slightly lower than the accuracy when using the original 3D modality-based features but with a 75% reduction ratio from using RGB modality. These results motivate implementing the proposed recognition model on an embedded system platform in the future.

• 전자공학회논문지B
• /
• 제33B권7호
• /
• pp.138-146
• /
• 1996

In this paper, w epropose a new mehtod to represent the shape and to recognize the object. The shape description and the matching is implemented by using the distance transform on the morphological skeleton. The employed distance transform is the chamfer (3,4) distance transform, because the chamfer distance transform (CDT) has an approximate value to the euclidean distance. The 2-D object can be represented by means of the distribution of the distance transform on the morphological skeleton, the number of skeletons, the sum of the CDT, and the other features are employed as the mtching parameters. The matching method has the invariant features (rotation, translation, and scaling), and then the method is used effectively for recognizing the differently-posed objects and/or marks of the different shape and size.

• International Journal of Internet, Broadcasting and Communication
• /
• 제12권4호
• /
• pp.83-92
• /
• 2020

A recent study to determine the fall is focused on analyzing fall motions using a recurrent neural network (RNN), and uses a deep learning approach to get good results for detecting human poses in 2D from a mono color image. In this paper, we investigated the improved detection method to estimate the position of the head and shoulder key points and the acceleration of position change using the skeletal key points information extracted using PoseNet from the image obtained from the 2D RGB low-cost camera, and to increase the accuracy of the fall judgment. In particular, we propose a fall detection method based on the characteristics of post-fall posture in the fall motion analysis method and on the velocity of human body skeleton key points change as well as the ratio change of body bounding box's width and height. The public data set was used to extract human skeletal features and to train deep learning, GRU, and as a result of an experiment to find a feature extraction method that can achieve high classification accuracy, the proposed method showed a 99.8% success rate in detecting falls more effectively than the conventional primitive skeletal data use method.

• ETRI Journal
• /
• 제44권2호
• /
• pp.286-299
• /
• 2022

Human activity recognition in real time is a challenging task. Recently, a plethora of studies has been proposed using deep learning architectures. The implementation of these architectures requires the high computing power of the machine and a massive database. However, handcrafted features-based machine learning models need less computing power and very accurate where features are effectively extracted. In this study, we propose a handcrafted model based on three-dimensional sequential skeleton data. The human body skeleton movement over a frame is computed through joint positions in a frame. The joints of these skeletal frames are projected into two-dimensional space, forming a "movement polygon." These polygons are further transformed into a one-dimensional space by computing amplitudes at different angles from the centroid of polygons. The feature vector is formed by the sampling of these amplitudes at different angles. The performance of the algorithm is evaluated using a support vector machine on four public datasets: MSR Action3D, Berkeley MHAD, TST Fall Detection, and NTU-RGB+D, and the highest accuracies achieved on these datasets are 94.13%, 93.34%, 95.7%, and 86.8%, respectively. These accuracies are compared with similar state-of-the-art and show superior performance.

• 한국산학기술학회논문지
• /
• 제22권2호
• /
• pp.127-133
• /
• 2021

낙상 판단을 위한 최근 발표되는 연구는 RNN(Recurrent Neural Network)을 이용한 낙상 동작 특징 분석과 동작 분류에 집중되어 있다. 웨어러블 센서를 기반으로 한 접근 방식은 높은 탐지율을 제공하나 사용자의 착용 불편으로 보편화 되지 못했고 최근 영상이나 이미지 기반에 딥러닝 접근방식을 이용한 낙상 감지방법이 소개 되었다. 본 논문은 2D RGB 저가 카메라에서 얻은 영상을 PoseNet을 이용해 추출한 인체 골격 키포인트(Keypoints) 정보로 머리와 어깨의 키포인트들의 위치와 위치 변화 가속도를 추정함으로써 낙상 판단의 정확도를 높이기 위한 감지 방법을 연구하였다. 특히 낙상 후 자세 특징 추출을 기반으로 Convolutional Neural Networks 중 Gated Recurrent Unit 기법을 사용하는 비전 기반 낙상 감지 솔루션을 제안한다. 인체 골격 특징 추출을 위해 공개 데이터 세트를 사용하였고, 동작분류 정확도를 높이는 기법으로 코, 좌우 눈 그리고 양쪽 귀를 포함하는 머리와 어깨를 하나의 세그먼트로 하는 특징 추출 방법을 적용해, 세그먼트의 하강 속도와 17개의 인체 골격 키포인트가 구성하는 바운딩 박스(Bounding Box)의 높이 대 폭의 비율을 융합하여 실험을 하였다. 제안한 방법은 기존 원시골격 데이터 사용 기법보다 낙상 탐지에 보다 효과적이며 실험환경에서 약 99.8%의 성공률을 보였다.