Robust Video Surveillance for Fall Detection Based on Human Shape Deformation

Rougier, C.; Meunier, Jean; St-Arnaud, Alain; Rousseau, Jacqueline

doi:10.1109/tcsvt.2011.2129370

Cited by 428 publications

(214 citation statements)

References 21 publications

Supporting

Mentioning

208

Contrasting

Unclassified

Order By: Relevance

“…With this method, they reported sensitivity and specificity values up to 99.70%, while we obtained 99.00% and 96.00%, respectively. Rougier et al [19], with an accuracy value of 99.70% compared to our 97.00%, performed fall detection in a video-level, not by stacks of frames. Hence, their system needs a specific data framing, using videos as input instead of a continuous stream of images, which is not ideal for real world deployments.…”

Section: Multicammentioning

confidence: 52%

“…To compare the best models found in the previous section with the state of the art, we used a 5-fold cross-validation for URFD and FDD and a leave-one-out cross-validation for Multicam following Rougier et al [19], in order to compare on equal conditions. In this last case, we split the dataset into 8 parts of the same size, each one containing all the videos recorded by a specific camera.…”

Section: Results and Comparison With The State Of The Artmentioning

confidence: 99%

“…Combining this information they are able to detect positions and also falls. Rougier et al [19] suggested using silhouettes as well, which is a common strategy in the literature. Applying a matching system along the video to track the deformation of the silhouette, they analyzed the shape of the body and finally obtained a result with a GMM.…”

Section: Related Workmentioning

confidence: 99%

“…(ii) Regarding the work by Auvinet et al [25] and Rougier et al [19], they are not included in the comparison tables due to their evaluation methodology (http://www.iro.umontreal.ca/∼labimage/Dataset/ technicalReport.pdf). They used a criterion that sets a variable called fall , representing the frame where the fall starts, therefore dividing each video into two parts.…”

Section: Multicammentioning

confidence: 99%

See 3 more Smart Citations

Vision-Based Fall Detection with Convolutional Neural Networks

Núñez-Marcos

Azkune

Arganda‐Carreras

2017

Wireless Communications and Mobile Computing

221

125

View full text Add to dashboard Cite

One of the biggest challenges in modern societies is the improvement of healthy aging and the support to older persons in their daily activities. In particular, given its social and economic impact, the automatic detection of falls has attracted considerable attention in the computer vision and pattern recognition communities. Although the approaches based on wearable sensors have provided high detection rates, some of the potential users are reluctant to wear them and thus their use is not yet normalized. As a consequence, alternative approaches such as vision-based methods have emerged. We firmly believe that the irruption of the Smart Environments and the Internet of Things paradigms, together with the increasing number of cameras in our daily environment, forms an optimal context for vision-based systems. Consequently, here we propose a vision-based solution using Convolutional Neural Networks to decide if a sequence of frames contains a person falling. To model the video motion and make the system scenario independent, we use optical flow images as input to the networks followed by a novel three-step training phase. Furthermore, our method is evaluated in three public datasets achieving the state-of-the-art results in all three of them.

show abstract

Section: Multicammentioning

confidence: 52%

Section: Results and Comparison With The State Of The Artmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Multicammentioning

confidence: 99%

See 2 more Smart Citations

Vision-Based Fall Detection with Convolutional Neural Networks

Núñez-Marcos

Azkune

Arganda‐Carreras

2017

Wireless Communications and Mobile Computing

221

125

View full text Add to dashboard Cite

show abstract

“…Since falls can have a devastating effect on the overall health and quality of living, many researches have actively been undertaken to monitor falls and activities of daily living (ADL) including walking patterns and postural changes. Several falls detection methods, using camera and video [2][3][4], tri-axial accelerometer [5,6], and gyroscope [7,8], have been studied to monitor falls. Among them, the tri-axial accelerometers and bi-axial gyroscope have been extensively used since these sensors have a good performance with high accuracy and reproducibility despite a relatively-low price.…”

Section: Introductionmentioning

confidence: 99%

Recognition of Falls and Activities of Daily Living using Tri-axial Accelerometer and Bi-axial Gyroscope

Park¹,

Kim²,

Kim³

et al. 2016

Journal of Sensor Science and Technology

View full text Add to dashboard Cite

This paper proposes a threshold-based fall recognition algorithm to discriminate between falls and activities of daily living (ADL) using a tri-axial accelerometer and a bi-axial gyroscope sensor mounted on the upper sternum. The experiment was executed ten times according to the proposed experimental protocol. The output signals of the tri-axial accelerometer and the bi-axial gyroscope were measured during eight falls and eleven ADL action sequences. The threshold values of the signal vector magnitude (SVM_Acc), angular velocity (ω res ), and angular variation (θ res ) parameter were calculated using MATLAB. From the preliminary study, three thresholds (TH1, TH2, and TH3) were set so that the falls could be distinguished from ADL. When the parameter SVM_Acc is greater than 2.5 g (TH1), ω res is greater than 1.75 rad/s (TH2), and θ res is greater than 0.385 rad (TH3), these action sequences are recognized as falls. If at least one or more of these conditions is not satisfied, the sequence is classified as ADL.

show abstract

Wear‐free indoor fall detection based on RFID and deep residual networks

Zhao

Zhu

et al. 2023

Int J Communication

View full text Add to dashboard Cite

Summary Since falls of the elderly can easily cause serious health problems in daily life, fall detection has received the attention of researchers. Traditionally, wearable sensors have been used to detect whether a person has fallen. However, wearable sensors may bring inconvenience to users' activities and affect user experience. In this paper, a fall detection approach based on RFID is proposed. In the proposed approach, non‐contact passive tags are used to construct an array of tags. Fall detection will be performed without the user wearing the device. The RSSI and phase data are collected when the reader queries the tags. Furthermore, an action segmentation algorithm is designed to quickly extract human action information based on the short‐term variance change of the phase signal. Subsequently, a deep residual network is built to classify fall movements and daily movements. Experiments show that the system can handle differences among users and locations and has an excellent performance in terms of recognition accuracy and efficiency, with an average accuracy rate of 96.77%.

show abstract

Robust Video Surveillance for Fall Detection Based on Human Shape Deformation

Cited by 428 publications

References 21 publications

Vision-Based Fall Detection with Convolutional Neural Networks

Vision-Based Fall Detection with Convolutional Neural Networks

Recognition of Falls and Activities of Daily Living using Tri-axial Accelerometer and Bi-axial Gyroscope

Wear‐free indoor fall detection based on RFID and deep residual networks

Contact Info

Product

Resources

About