Conventional Video Camera Research Articles

CamDec: Advancing Axis P1435-LE video camera security using honeypot-based deception

The explosion of online video streaming in recent years resulted in advanced services both in terms of efficiency and convenience. However, Internet-connected video cameras are prone to exploitation, leading to information security issues and data privacy concerns. The proliferation of video-capable Internet of Things devices and cloud-managed surveillance systems further extend these security issues and concerns. In this paper, a novel approach is proposed for video camera deception via honeypots, offering increased security measures compared to what is available on conventional Internet-enabled video cameras.

Heart rate estimation from color video sequences with high SNR.

We propose an absorption intensity heartbeat modulation-averaged shifted histogram (AIHM-ASH) method for estimating human heart rate (HR) using color videos of lip image sequences. When heartbeat occurs, AIHM is generated. Based on the AIHM, HR signals can be demodulated by computing the instantaneous HR modulation depth that presents the relative red blood cell (RBC) concentration from the green channel image of the RGB color video. In addition, the ASH algorithm further suppresses the background tissue and vein signals, and increases the signal-to-noise ratio (SNR). The experimental results for flow phantoms, chicken embryos, and human lips validated the proposed method's optimal estimation conditions and effectiveness, where the accuracy and root mean square error (RMSE) were 99.23% and 0.8 bpm, respectively. The proposed HR estimation method has significant potential to advance health monitoring and disease prevention via conventional color video cameras installed in public places.

Automatic Assessment of the Type and Intensity of Agitated Hand Movements

With increasing numbers of people living with dementia, there is growing interest in the automatic monitoring of agitation. Current assessments rely on carer observations within a framework of behavioural scales. Automatic monitoring of agitation can supplement existing assessments, providing carers and clinicians with a greater understanding of the causes and extent of agitation. Despite agitation frequently manifesting in repetitive hand movements, the automatic assessment of repetitive hand movements remains a sparsely researched field. Monitoring hand movements is problematic due to the subtle differences between different types of hand movements and variations in how they can be carried out; the lack of training data creates additional challenges. This paper proposes a novel approach to assess the type and intensity of repetitive hand movements using skeletal model data derived from video. We introduce a video-based dataset of five repetitive hand movements symptomatic of agitation. Using skeletal keypoint locations extracted from video, we demonstrate a system to recognise repetitive hand movements using discriminative poses. By first learning characteristics of the movement, our system can accurately identify changes in the intensity of repetitive movements. Wide inter-subject variation in agitated behaviours suggests the benefit of personalising the recognition model with some end-user information. Our results suggest that data captured using a single conventional RGB video camera can be used to automatically monitor agitated hand movements of sedentary patients.

Hold that pose: capturing cervical dystonia's head deviation severity from video

ObjectiveDeviated head posture is a defining characteristic of cervical dystonia (CD). Head posture severity is typically quantified with clinical rating scales such as the Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS). Because clinical rating scales are inherently subjective, they are susceptible to variability that reduces their sensitivity as outcome measures. The variability could be circumvented with methods to measure CD head posture objectively. However, previously used objective methods require specialized equipment and have been limited to studies with a small number of cases. The objective of this study was to evaluate a novel software system—the Computational Motor Objective Rater (CMOR)—to quantify multi‐axis directionality and severity of head posture in CD using only conventional video camera recordings.MethodsCMOR is based on computer vision and machine learning technology that captures 3D head angle from video. We used CMOR to quantify the axial patterns and severity of predominant head posture in a retrospective, cross‐sectional study of 185 patients with isolated CD recruited from 10 sites in the Dystonia Coalition.ResultsThe predominant head posture involved more than one axis in 80.5% of patients and all three axes in 44.4%. CMOR's metrics for head posture severity correlated with severity ratings from movement disorders neurologists using both the TWSTRS‐2 and an adapted version of the Global Dystonia Rating Scale (rho = 0.59–0.68, all p <0.001).ConclusionsCMOR's convergent validity with clinical rating scales and reliance upon only conventional video recordings supports its future potential for large scale multisite clinical trials.

Determination of the Optimum Excitation Wavelength for the Parathyroid Gland Using a Near-Infrared Camera.

When performing thyroid/parathyroid surgery, difficulty detecting the parathyroid gland is a common experience because it is frequently mistaken with surrounding structures, including the thyroid gland, lymph nodes, and fat. To obtain successful surgical results, the auto fluorescent property of the parathyroid gland occurring at 820–830 nm has been used. Intraoperative visualization and detection by fluorescence enable protection of the gland from damage and unintended removal. Use of a near-infrared (NIR) camera has been proposed to indicate the parathyroid gland, but the devices and success rates have varied. This study aimed to define optimum excitation wavelength (EWL) by measuring the EWL of the parathyroid gland for its autofluorescence. Glands were exposed to EWL at 10-nm intervals from 670–790 nm with a light-emitting diode monochromator; autofluorescence intensity was recorded with a conventional NIR video camera. Autofluorescence intensity curves of three normal parathyroid glands were depicted; the optimum EWL was measured as 760–770 nm. Also, the illumination of the surrounding structures were compared at the optimum EWL. The auto fluorescent intensity of the parathyroid gland was 2-fold greater than for surrounding structures. This difference in fluorescence intensity should enable distinction of the parathyroid gland from surrounding structures. The clarification of the optimum EWL can guide refinements of the NIR camera for better surgical outcomes by improving detection of the parathyroid glands. Also, an understanding of optimum EWL should lead to developments for microscopic devices to unravel the still unknown mechanisms of the intrinsic autofluorescence of the parathyroid gland.

Investigating the significance of adversarial attacks and their relation to interpretability for radar-based human activity recognition systems

Given their substantial success in addressing a wide range of computer vision challenges, Convolutional Neural Networks (CNNs) are increasingly being used in smart home applications, with many of these applications relying on the automatic recognition of human activities. In this context, low-power radar devices have recently gained in popularity as recording sensors, given that the usage of these devices allows mitigating a number of privacy concerns, a key issue when making use of conventional video cameras. Another concern that is often cited when designing smart home applications is the resilience of these applications against cyberattacks. It is, for instance, well-known that the combination of images and CNNs is vulnerable against adversarial examples, mischievous data points that force machine learning models to generate wrong classifications during testing time. In this paper, we investigate the vulnerability of radar-based CNNs to adversarial attacks, and where these radar-based CNNs have been designed to recognize human gestures. Through experiments with four unique threat models, we show that radar-based CNNs are susceptible to both white- and black-box adversarial attacks. We also expose the existence of an extreme adversarial attack case, where it is possible to change the prediction made by the radar-based CNNs by only perturbing the padding of the inputs, without touching the frames where the action itself occurs. Moreover, we observe that gradient-based attacks exercise perturbation not randomly, but on important features of the input data. We highlight these important features by making use of Grad-CAM, a popular neural network interpretability method, hereby showing the connection between adversarial perturbation and prediction interpretability.

МЕТОДИКА ОСВОЕНИЯ БИОМЕХАНИЧЕСКОЙ СТРУКТУРЫ СОСКОКА «ДВОЙНОЕ САЛЬТО ВПЕРЕД В ГРУППИРОВКЕ НА ПАРАЛЛЕЛЬНЫХ БРУСЬЯХ МАХОМ НАЗАД ИЗ УПОРА»

Statement of the problem. In exercises on parallel bars, the double somersault forward dismount in a tuck position with a backward swing from the stop is the culminating final chord that completes the logical, aesthetically solid construction created by the gymnast’s motor means. Currently, this is one of the most difficult modern jumps, often performed by highly qualified gymnasts at major competitions. Landing on the board after it at the end of the combination significantly increases the gymnast’s chances of getting a high score. However, the phasal biomechanical structure of its execution technique has not been studied enough. This is due to the use of conventional video cameras with a low shooting speed (25 fps). The problem is to develop a method for mastering the technique of performing the double somersault forward dismount in a tuck position on parallel bars swinging backward from the stop using a high-speed Phantom camcorder at a speed of 500 fps. The purpose of the article is to identify the phasal biomechanical structure of the technique of performing a double somersault forward dismount in a tuck position on parallel bars with a backward swing from the stop and to develop a method of its mastering. The research methodology consists of a complex of instrumental methods, in particular using high-speed video camera Phantom at a speed of 500 fps, software “Tema MaxTrack” to process video materials, computer technology, and theoretical analysis of literature of international and Russian scientists and policy documents, teacher observations, structural-phase and pedagogical and biomechanical analysis. Research results. The phasal biomechanical structure of the technique of performing a gymnastic element – a double somersault forward dismount in a tuck position on parallel bars – is revealed. An effective method of teaching this element has been developed and justified. Conclusion. Using modern instrumental methods, the biomechanical phase structure of the technique of performing a double somersault forward dismount in a tuck position on parallel bars was studied on the basis of the methodology of the interdisciplinary direction of pedagogical biomechanics. The biomechanical characteristics of the phase microstructure were determined using data from high-speed video (500 fps) and a pedagogical-biomechanical structural-phase model was constructed.

A Method for Real-Time Patient Positioning Verification in Radiotherapy Using Multi-Angular Images with Conventional Video Cameras

Background: Accurate patient positioning and the corresponding position verification play a key role in precision radiotherapy. The conventional position verification techniques, such as cone-beam computed tomography (CBCT), used in radiotherapy deliver additional radiation doses to patients; thus some complications may occur in the surrounding normal tissue. Additionally, the present registration and fusion algorithms require a long time to verify and monitor the patient's position in real-time. Method: A new simple method using three conventional video cameras from different directions for observing the changes in the patient's position and taking the multiangular surface images of the patient without delivering additional doses of radiation is proposed in this paper. Real-time patient position verification could be achieved through this Oriented FAST and Rotated BRIEF (ORB) algorithm modified by us. Results: A plastic patient model immobilized with thermoplastic shell on a simulation couch was used to verify the accuracy and feasibility of our method. Testing results showed that the position and angle offset errors identified with our system were less than 1 mm and 1°, respectively. Moreover, the modified ORB algorithm was fast enough for real-time images comparison. Conclusion: This proposed method shows a great potential for application in verifying and monitoring the patient's position in real time during radiotherapy.

Blinkies: Open Source Sound-to-Light Conversion Sensors for Large-Scale Acoustic Sensing and Applications

We propose an open source hardware platform called Blinky, a sound-to-light conversion sensor that harvests sound power at low-rate and conveniently. Blinkies are made up of a central processing unit connected to two microphones and a few light-emitting devices, are powered by a battery, and protected by a robust enclosure. Distributed in space and combined with a conventional video camera, they allow to practically sense sound power over a very large area without hassle. We give a comprehensive overview of the proposed system and its potential applications. We describe the hardware design and trade-offs made. We provide a model for the channel between sound power measurements and signal acquired by the video camera. Because each sensor is potentially affected by a different attenuation due to the channel, we propose a calibration procedure to restore the scale of the measurements. The effectiveness of the calibration is validated in an experiment. Finally, we demonstrate sound source localization using a hundred-and-one actual Blinkies in highly reverberent environment.

Fast surveillance video indexing &amp; retrieval with WiFi MAC address tagging

&lt;span&gt;Conventional public safety surveillance video camera systems required 24/7 monitoring of security officers with video wall display installed in the control room. When a crime or incident is reported, all the recorded surveillance video streams nearby the incident area are playback simultaneously on video wall to help locate the target person. The security officers can fast forward the video playback to speed up the video search, but it requires massive manpower if there are hundreds of video streams required to be examined on the video wall. One of the possible solutions is through a suitable video indexing and retrieval technique to prioritize the video frames that need to be processed. This paper presents a WiFi sniffer enabled surveillance camera, with 3-stage WiFi frame inspection filter and the use of collected WiFi signal strength for filtering, to tag the collected WiFi MAC addresses to the surveillance video frames according to the time of the MAC address is sniffed. Additional metadata (WiFi MAC address of smartphone) collected during the occurrence of the incident can be used to prioritize the retrieving of surveillance video frames for subsequent image processing. &lt;/span&gt;

Automation of the Timed-Up-and-Go Test Using a Conventional Video Camera.

The Timed-Up-and-Go (TUG) test is a simple clinical tool commonly used to quickly assess the mobility of patients. Researchers have endeavored to automate the test using sensors or motion tracking systems to improve its accuracy and to extract more resolved information about its sub-phases. While some approaches have shown promise, they often require the donning of sensors or the use of specialized hardware, such as the now discontinued Microsoft Kinect, which combines video information with depth sensors (RGBD). In this work, we leverage recent advances in computer vision to automate the TUG test using a regular RGB video camera without the need for custom hardware or additional depth sensors. Thirty healthy participants were recorded using a Kinect V2 and a standard video feed while performing multiple trials of 3 and 1.5 meter versions of the TUG test. A Mask Regional Convolutional Neural Net (R-CNN) algorithm and a Deep Multitask Architecture for Human Sensing (DMHS) were then used together to extract global 3D poses of the participants. The timing of transitions between the six key movement phases of the TUG test were then extracted using heuristic features extracted from the time series of these 3D poses. The proposed video-based vTUG system yielded the same error as the standard Kinect-based system for all six key transitions points, and average errors of less than 0.15 seconds from a multi-observer hand labeled ground truth. This work describes a novel method of video-based automation of the TUG test using a single standard camera, removing the need for specialized equipment and facilitating the extraction of additional meaningful information for clinical use.

普通摄像机图像引导的放射治疗技术

普通摄像机图像引导的放射治疗技术

Laser and Optical Flow Fusion for a Non-Intrusive Obstacle Detection System on an Intelligent Wheelchair

In this paper, a method to combine a low-cost 16 beam solid state laser sensor and a conventional video camera for obstacle detection is presented. The system is intended to form a non-intrusive virtual barrier at both sides of an intelligent wheelchair, in order to protect the user in everyday outdoor and indoor environments, like office, home, or pedestrian areas. In this type of environments, the shape of the obstacles is very heterogeneous (for instance, having a variable width along their height), so their detection via the conventional sensors installed in the wheelchair presents difficulties, especially in the most exposed areas of the user that are the sides of their torso. With the proposed system, when an obstacle intersects a beam, the intersection point is projected in the associated image and optical flow is calculated at both sides of this point. Using the optical flow, several classifiers have been trained and tested in order to automatically discern between intersections produced by the user and those produced by external obstacles. Results show that the method accurately detects the presence of obstacles and the direction of their movement.

Transdermal Optical Imaging Reveal Basal Stress via Heart Rate Variability Analysis: A Novel Methodology Comparable to Electrocardiography.

The present study examined the validity of a novel physiological measurement technology called transdermal optical imaging (TOI) technology at assessing basal stress. This technology conveniently, contactlessly, and remotely measures facial blood flow changes using a conventional digital video camera. We compared data from TOI against the pulse data collected from the FDA approved BIOPAC system. One hundred thirty-six healthy adults participated in the study. We found that TOI measurements of heart rate and heart rate variability (HRV), which reflects basal stress, corresponded strongly to those obtained from BIOPAC. These findings indicate that TOI technology is a viable method to monitor heart rate and HRV not only accurately but also conveniently, contactlessly, and remotely. Further, measures of HRV obtained via TOI serves as a valid index of basal stress. Potential applications of this technology in psychological research and other fields are discussed.

Video-Based Outdoor Human Reconstruction

A human body scanning system of great practical convenience, which can be used in an outdoor environment, is proposed. The system uses only a single conventional video camera without the aid of special sensors or controlled illuminations. We leverage the structure from motion calibration results directly and improve the available video-based dense 3D reconstruction by integrating the surface smoothness constraints. The point cloud reinforcement is proposed to detect and adjust the conflict point data for the slender and shaky body parts. Combined with the silhouette adaptation, the proposed point cloud reinforcement achieves reasonable and plausible mesh reconstruction on these challenging parts. We further introduce the close-shot frames to refine the prereconstructed mesh model, leading to a colored watertight model. The overall system is approximate to automatic since only one or two times of painting brush interaction are required for robust and high-quality multiview image segmentation. The experiment results on various test sequences demonstrate the effectiveness and the robustness of the proposed method, even under very challenging scenarios when shaking body, varying illumination, and textureless regions occur.

Potential Utility of a 4K Consumer Camera for Surgical Education in Ophthalmology.

Purpose. We evaluated the potential utility of a cost-effective 4K consumer video system for surgical education in ophthalmology. Setting. Tokai University Hachioji Hospital, Tokyo, Japan. Design. Experimental study. Methods. The eyes that underwent cataract surgery, glaucoma surgery, vitreoretinal surgery, or oculoplastic surgery between February 2016 and April 2016 were recorded with 17.2 million pixels using a high-definition digital video camera (LUMIX DMC-GH4, Panasonic, Japan) and with 0.41 million pixels using a conventional analog video camera (MKC-501, Ikegami, Japan). Motion pictures of two cases for each surgery type were evaluated and classified as having poor, normal, or excellent visibility. Results. The 4K video system was easily installed by reading the instructions without technical expertise. The details of the surgical picture in the 4K system were highly improved over those of the conventional pictures, and the visual effects for surgical education were significantly improved. Motion pictures were stored for approximately 11 h with 512 GB SD memory. The total price of this system was USD 8000, which is a very low price compared with a commercial system. Conclusion. This 4K consumer camera was able to record and play back with high-definition surgical field visibility on the 4K monitor and is a low-cost, high-performing alternative for surgical facilities.

Sensing Method of Patient’s Body Movement Without Attaching Sensors on the Patient’s Body -Evaluation of “Scratching Cheek”, “Turning Over and Scratching Back” and “Scratching Shin”-

Medical specialists hardly pay attention to scratching as a disease, and thus, few measurement methods of scratching behavior and frequencies have been considered. Scratching can yield serious mental and physical problems. This paper describes a monitoring method for scratching and turning over motions during sleep, using a sensor without a power supply. The method is based on the piezoelectric effect. Piezo-ceramic devices are placed under the legs of a bed. When a subject lies on the bed, body movements occur. Small body movements can occur because of heart beats, while breathing, scratching, and other physiological phenomena. Significant body movements occur while changing positions during sleeping or while moving out of the bed. The proposed method detects these body movements, including the scratching. Here, we aimed at investigating how clearly the proposed method discriminates scratching from the other motions and extracts the scratching period. To validate the proposed method, we carried out experiments and evaluated three cases involving the scratching motion. In Case 1, the subject scratches one’s right cheek with right hand; in Case 2, the subject scratches one’s back after turning over; in Case 3, the subject scratches one’s left shin with right foot. The results showed that the proposed method, using this simple device, can detect scratching, turning over, and their occurring periods with a comparable degree of accuracy to the conventional video camera recording method. Furthermore, the sensor could record a signal during the night, under real conditions, when a subject sleeps on the equipped bed.

A Robust Visual-Feature-Extraction Method for Simultaneous Localization and Mapping in Public Outdoor Environment

We describe a new feature extraction method based on the geometric structure of matched local feature points that extracts robust features from an image sequence and performs satisfactorily in highly dynamic environments. Our proposed method is more accurate than other such methods in appearance-only simultaneous localization and mapping (SLAM). Compared to position-invariant robust features [1], it is also more suitable for low-cost, single lens cameras with narrow fields of view. Testing our method in an outdoor environment at Shibuya Station. We captured images using a conventional hand-held single-lens video camera. These environments of experiments are public environments without any planned landmarks. Results have shown that the proposed method accurately obtains matches for two visual-feature sets and that stable, accurate SLAM is achieved in dynamic public environments.

A Prototype System for Atmospheric Visibility Estimation Based on Image Analysis and Learning

Visibility plays an important role in transportation, communication, agriculture, industry and so on. It is also an indication of air quality. This paper presents the design, implementation, and evaluation of a prototype system for estimating visibility quantitatively based on image analysis and learning. Firstly, an image of measured scene including objects and their background was captured by conventional video camera and the image features were extracted from spatial and frequency domain respectively. Secondly, SVR (Support Vector Regression) model was dynamically produced with the training set according to the image. Lastly, visibility of the image was calculated by SVR model with its features as input vector. A ground truth dataset covering the whole range is used to evaluate the performance of the system and the experimental results show that the system has both simple operation and high accuracy, and it is stable enough to be applied for the whole range of visibility in practical applications.

Geometric and radiometric characteristics of a prototype surveillance system

This paper reports the research findings regarding the quality of the captured video of a prototype Near-Infrared imaging system which was developed for night-time surveillance. The prototype required the incorporation of a 3D optical stereo adapter and a set of external near-infrared filters onto the optics of a conventional video camera. The study found that the video captured by the prototype was approximately 85% of the original video camera resolution and the radial lens distortion increased from one to approximately two pixels. The distortion was easily reduced to sub-pixel value using suitable image processing techniques. The standard deviation of the 3D measurement between any two facial anthropometric marks was better than 2mm for all the tested models. The image radiometric quality was sufficient to provide accurate facial landmark identification for three test subjects. The results confirmed that the prototype was suitable for 3D spatial data capture and feature interpretation of crime suspects in court.