An ultra-fast human detection method for color-depth camera

Li, Jun; Zhang, Guyue; Liu, Ye; Tian, Luchao; Chen, Yan Qiu

doi:10.1016/j.jvcir.2015.06.014

Cited by 26 publications

(23 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other classical learning approaches have also been proposed for identifying human body parts in outdoor environments, such as parking lots and town centers [17], and indoor environments, such as retail stores and offices [18]. For example, in [18], a two-stage procedure was used for detecting the top of human heads using RGB and depth.…”

Section: Person Detection Using Classical Machine Learning Approachesmentioning

confidence: 99%

“…In [5], a long short-term memory (LSTM) network was used to detect head-tops. The first layer employed the headtop detection technique presented in [18], where for each possible head-top pixel, a set of bounding boxes were generated from both RGB and depth images. This set of boxes contained different ratios of potential human body proportions for a particular head-top.…”

Section: Person Detection Using Deep Learning Approaches With Rgb Andmentioning

confidence: 99%

See 1 more Smart Citation

Using Deep Learning to Find Victims in Unknown Cluttered Urban Search and Rescue Environments

et al. 2020

View full text Add to dashboard Cite

Purpose of Review We investigate the first use of deep networks for victim identification in Urban Search and Rescue (USAR). Moreover, we provide the first experimental comparison of single-stage and two-stage networks for body part detection, for cases of partial occlusions and varying illumination, on a RGB-D dataset obtained by a mobile robot navigating cluttered USAR-like environments. Recent Findings We considered the single-stage detectors Single Shot Multi-box Detector, You Only Look Once, and RetinaNet and the two-stage Feature Pyramid Network detector. Experimental results show that RetinaNet has the highest mean average precision (77.66%) and recall (86.98%) for detecting victims with body part occlusions in different lighting conditions. Summary End-to-end deep networks can be used for finding victims in USAR by autonomously extracting RGB-D image features from sensory data. We show that RetinaNet using RGB-D is robust to body part occlusions and low-lighting conditions and outperforms other detectors regardless of the image input type.

show abstract

Section: Person Detection Using Classical Machine Learning Approachesmentioning

confidence: 99%

Section: Person Detection Using Deep Learning Approaches With Rgb Andmentioning

confidence: 99%

Using Deep Learning to Find Victims in Unknown Cluttered Urban Search and Rescue Environments

et al. 2020

View full text Add to dashboard Cite

show abstract

“…1(a) shows a typical result of this method. Readers are referred to [20] for more details of the proposal generation method.…”

Section: Proposal Generationmentioning

confidence: 99%

“…The set of images includes a headscale, an upperbody-scale, a body-scale and several larger ones. Clipping sizes of head and body are obtained with their real-world sizes, depth value of the plausible head-top pixel and depth camera's intrinsic parameters [20]. Size of nth peripheral image is calculated as: S n = S b * (1 + 0.3n) in which S b denotes the size of body scale image.…”

Section: Multi-glimplse Lstmmentioning

confidence: 99%

Multi-glimpse LSTM with color-depth feature fusion for human detection

Liu

Zhang

et al. 2017

2017 IEEE International Conference on Image Processing (ICIP)

Self Cite

View full text Add to dashboard Cite

With the development of depth cameras such as Kinect and Intel Realsense, RGB-D based human detection receives continuous research attention due to its usage in a variety of applications. In this paper, we propose a new Multi-Glimpse LSTM (MG-LSTM) network, in which multi-scale contextual information is sequentially integrated to promote the human detection performance. Furthermore, we propose a feature fusion strategy based on our MG-LSTM network to better incorporate the RGB and depth information. To the best of our knowledge, this is the first attempt to utilize LSTM structure for RGB-D based human detection. Our method achieves superior performance on two publicly available datasets.

show abstract

“…Ghost Sequences It is unavoidable for a video camera tracking body motion to report bounding boxes that do not correspond to an actual user or object, despite efforts in human detection research [9]. Common causes include lingering images that persist after the user has moved or vibrations applied directly or indirectly to the camera.…”

Section: Video-based Locationmentioning

confidence: 99%

Rule-Based Real-Time ADL Recognition in a Smart Home Environment

Baryannis

Woznowski

Antoniou

2016

Rule Technologies. Research, Tools, and Applications

View full text Add to dashboard Cite

General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Abstract. This paper presents a rule-based approach for both offline and real-time recognition of Activities of Daily Living (ADL), leveraging events produced by a non-intrusive multi-modal sensor infrastructure deployed in a residential environment. Novel aspects of the approach include: the ability to recognise arbitrary scenarios of complex activities using bottom-up multi-level reasoning, starting from sensor events at the lowest level; an effective heuristics-based method for distinguishing between actual and ghost images in video data; and a highly accurate indoor localisation approach that fuses different sources of location information. The proposed approach is implemented as a rule-based system using Jess and is evaluated using data collected in a smart home environment. Experimental results show high levels of accuracy and performance, proving the effectiveness of the approach in real world setups.

show abstract

An ultra-fast human detection method for color-depth camera

Cited by 26 publications

References 29 publications

Using Deep Learning to Find Victims in Unknown Cluttered Urban Search and Rescue Environments

Using Deep Learning to Find Victims in Unknown Cluttered Urban Search and Rescue Environments

Multi-glimpse LSTM with color-depth feature fusion for human detection

Rule-Based Real-Time ADL Recognition in a Smart Home Environment

Contact Info

Product

Resources

About