Abstract
Falls have caught great harm to the elderly living alone at home. This paper presents a novel visual-based fall detection approach by Dual-Channel Feature Integration. The proposed approach divides the fall event into two parts: falling-state and fallen-state, which describes the fall events from dynamic and static perspectives. Firstly, the object detection model (Yolo) and the human posture detection model (OpenPose) are used for preprocessing to obtain key points and the position information of a human body. Then, a dual-channel sliding window model is designed to extract the dynamic features of the human body (centroid speed, upper limb velocity) and static features (human external ellipse). After that, MLP (Multilayer Perceptron) and Random Forest are applied to classify the dynamic and static feature data separately. Finally, the classification results are combined for fall detection. Experimental results show that the proposed approach achieves an accuracy of 97.33% and 96.91% when tested with UR Fall Detection Dataset and Le2i Fall Detection Dataset.
Highlights
With the development of society and the improvement of medical standards, the aging of the population has become a global trend
The rate of centroid drop, the speed of the upper limbs, the location of key points, and the ellipse parameters of the human body are computed. These features are further divided into types of dynamic and static, which are used for describing the human body during the fall event
DATA DESCRIPTION In order to prove the validity of our method, experiments were conducted on three public databases: Fall Detection Dataset, Le2i Fall Detection Dataset, and UR Fall Detection Dataset
Summary
With the development of society and the improvement of medical standards, the aging of the population has become a global trend. An Ambient device-based method detects the fall event through the sound, vibration, and other signals collected by sensors installed in the room (walls, floors, etc.). Montanini et al [8] proposed a detection approach based on footwear They installed a force sensor and a three-axis acceleration sensor in shoes to analyze the fall event. A Computer vision-based method generally detects fall events by analyzing the video or image. For the wearable sensor-based method, it has high accuracy, these devices need to be carried around, which adds uncertainty to the fall detection. Video-based fall detection requires additional equipment (cameras, embedded devices, etc.), this is tolerable compared to its advantages. Dual-channel sliding window model (DSW) is proposed to extract dynamic and static features.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
