Webcam Research Articles

SOFT: Self-supervised sparse Optical Flow Transformer for video stabilization via quaternion

Video stabilization is crucial for video representation learning, which suffers from the challenges such as the perception of unstable vision, the stripping and cognition of target motion features in complex scenes, the correction of the jittery camera systems trails. In this paper, we propose a Self-supervised sparse Optical Flow Transformer (SOFT) model, consisting of a self-supervised contrastive learning transformer network, a sparse optical flow perception network and a multimodal cognitive fusion network. The SOFT model takes advantage of optical flow to estimate motion. The sparse optical flow perception network perceiving partially sparse optical flow containing motion features. This serves as the input to the self-supervised contrastive learning transformer network for generating sparse optical flow features, which are fed into the multimodal cognitive fusion network together with the real and virtual camera pose for video frame warping. Experimental comparisons with state-of-the-art models on 4 metrics demonstrate the effectiveness of the SOFT model. It achieves the best performance with an average Stability of 0.869 and average Distortion of 0.993 across 6 categories videos, which shows that the SOFT model can effectively perceive the motion in the video and smooth the jitter track of videos.

Application of a stand-alone RTI measuring system with an integrated camera in cultural heritage digitisation

The Multi-Light Image Collection (MLIC) data, in most cases processed with the Reflectance Transformation Imaging (RTI) technique, is one of the popular methods of documentation and digital representation of movable finds from an archaeological excavation. The manual (hand-held) approach to collecting this data is technically simple and still popular in the case of archaeology. However, an extensive amount of work during image collection is a strong motivation for building RTI acquisition systems, which are also popular, in most cases, in the shape of domes. The research presented in the article focuses on two aspects of building such a device. First, the study examines the effect of the placement and number of lights on the resulting RTI model for a system based on LED stripe-shaped circuits. Secondly, the camera component was studied.Recent development in the area of low-cost accessories for microcomputers resulted in a few new industrial-grade USB cameras with the possibility to take high-resolution images. Cameras designed for use with microcomputers offer broad control of the focus or exposition parameters and provide the possibility to build RTI domes with the fully integrated camera, therefore, almost complete automatization of the capturing and processing of MLIC data.This paper presents the main assumptions and elements of the dome-shaped measurement system, which enables automatic RTI data acquisition and processing together with the evaluation of light configuration used to build relightable models. Amphoras handles with stamps and ancient pottery were used as an objects to evaluate the quality of the relightable models generated from the images acquired by a programmable miniature Arducam camera. Results of comparison with professional high-quality mirrorless cameras show that the quality of images from miniature cameras can be satisfactory for some archaeological objects.

Smart Face Attendance System

Smart Face Attendance System

Raspberry Pi based Web Monitoring Smart Security System

The surveillance system has been utilized over the years on public property in order to provide security and prevent unauthorized entrance. Lately, more private homeowners are choosing to implement security systems. There are many problems with the video surveillance system. These disadvantages are the indistinctiveness of the pictures/video and the need for a lot of storage space to save surveillance information. This paper describes the design and implementation of a low-cost Web Monitoring System based on Raspberry Pi. Also, a web application is designed for the purposes of controlling the camera and for live streaming. The live stream from a web camera can be viewed from any web browser, even mobile, in real-time. Controlling the camera is possible with the web application using four commands (up/down, left/right). Also, the camera moves depending on the detection of an object located in the visible area of the PIR Motion sensor, thereby issuing a warning message about object detection.

Cut2Self: A single image based self‐supervised denoiser

AbstractDespite the recent upsurge of self‐supervised methods in single image denoising, achieving robustness and efficiency of performance is still challenging due to some prevalent issues like identity mapping, overfitting, and increased variance of network predictions. Recent self‐supervised approaches prescribe a dropout‐based single‐pixel masking strategy in this regard. However, real camera noise is signal‐dependent, and typically poses trivial changes to the images. Hence, such a strategy still preserves contextual information about target location even after dropping them out, leading to an identity mapping and overfitting problems in practice. Here, Cut2Self, a new denoising method to address this issue, which cuts out random block‐regions instead of singleton pixels to provide the higher possibility to remove contextual information from the neighbouring pixels, thus reducing identity mapping chances while being resilient against overfitting is proposed. Cut2Self creates distinct training pairs for each training iteration by randomly cutting out square regions of input and sending them to the denoising network. Thus, iteration‐wise different network predictions are generated, which are then assembled to generate the final denoised output. Cut2Self is evaluated with synthetic and real‐world noise, visualising its consistent denoising performance compared to other supervised, unsupervised, and self‐supervised methods.

Implementation of a Web Camera System in an Australian Neonatal Intensive Care Unit: Pre- and Postevaluation of the Parent and Staff Experience.

Admission to a neonatal intensive care unit (NICU) for prematurity or illness is necessary for approximately 20% of newborns in Australia, resulting in parent-infant separation. Web cameras in the NICU provide a virtual link for parents to remain remotely connected to their infant during admission. Web camera use is increasing; however, there is limited evidence on the impact of web cameras on parents, infants, and neonatal staff. There were two objectives: (1) to determine the attitudes of parents and staff toward web cameras in the NICU and (2) to compare parental depression, anxiety, and stress levels using validated scales before and after web camera implementation in the NICU. A pre- and postevaluation survey was administered before and after implementation of the NICVIEW camera system in a tertiary NICU in Sydney, Australia. The NICVIEW camera system provides secure real-time viewing of infants and can be accessed from any device with an internet connection. Surveys were administered to parents of inpatients and staff, and included open- and closed-ended questions and Likert scales. Survey questions aimed to determine parent and staff attitudes and use of web cameras before and after implementation. In addition, pre- and postimplementation parental levels of depression, anxiety, and stress, as measured by the 21-item version of the Depression Anxiety Stress Scale (DASS-21) and Parental Stressor Scale: Neonatal Intensive Care Unit, were recorded. In total, 94 parents and 109 staff members completed the pre- and postimplementation surveys. Post implementation, 43 of 44 (98%) parents supported web cameras, and 40 of 42 (95%) parents stated that they used web cameras. The most common reasons for support from parents included web cameras making parents feel more at ease, facilitating parent-infant bonding, increasing parental confidence in staff, and allowing others to see infants. There was no significant difference between the parental groups for the depression, anxiety, or stress scales measured by DASS-21. Staff support for web cameras increased significantly from 34 of 42 (81%) participants before to 64 of 67 (96%) participants after implementation (P=.01). Following implementation, there was a resolution in staff concerns about web cameras having an adverse impact on staff roles and privacy and security concerns. Web camera use in a tertiary Australian NICU was strongly supported by parents and staff and may reduce parental stress, facilitate parent-infant bonding, and encourage positive parent-staff engagement. Web cameras are a feasible method of providing continuity of care for families and should be considered as a standard of care in similarly resourced settings.

Assessment of a Machine Learning Algorithm Using Web Images for Flood Detection and Water Level Estimates

Improving our skills to monitor flooding events is crucial for protecting populations and infrastructures and for planning mitigation and adaptation strategies. Despite recent advancements, hydrological models and remote sensing tools are not always useful for mapping flooding at the required spatial and temporal resolutions because of intrinsic model limitations and remote sensing data. In this regard, images collected by web cameras can be used to provide estimates of water levels during flooding or the presence/absence of water within a scene. Here, we report the results of an assessment of an algorithm which uses web camera images to estimate water levels and detect the presence of water during flooding events. The core of the algorithm is based on a combination of deep convolutional neural networks (D-CNNs) and image segmentation. We assessed the outputs of the algorithm in two ways: first, we compared estimates of time series of water levels obtained from the algorithm with those measured by collocated tide gauges and second, we performed a qualitative assessment of the algorithm to detect the presence of flooding from images obtained from the web under different illumination and weather conditions and with low spatial or spectral resolutions. The comparison between measured and camera-estimated water levels pointed to a coefficient of determination R2 of 0.84–0.87, a maximum absolute bias of 2.44–3.04 cm and a slope ranging between 1.089 and 1.103 in the two cases here considered. Our analysis of the histogram of the differences between gauge-measured and camera-estimated water levels indicated mean differences of −1.18 cm and 5.35 cm for the two gauges, respectively, with standard deviations ranging between 4.94 and 12.03 cm. Our analysis of the performances of the algorithm to detect water from images obtained from the web and containing scenes of areas before and after a flooding event shows that the accuracy of the algorithm exceeded ~90%, with the Intersection over Union (IoU) and the boundary F1 score (both used to assess the output of segmentation analysis) exceeding ~80% (IoU) and 70% (BF1).

BRAS: Bidirectional Reflectance Adjustment Strategy for 3-D Reconstruction of Mirror-Like Surface

A mirror-like surface (MLS) reflects highlight, aggravating the image saturation in structured light three-dimensional (3D) reconstruction systems and precluding defect detection based on reconstructed 3D profiles. Previous studies have focused on a strategy of limiting the luminous flux entering the camera. However, the high intensity of the reflected highlight force the limitation to be strengthened, which heavily reduces the modulation in the images for the 3D reconstruction. Therefore, we propose a new strategy to adjust the source of the reflected highlight, i.e., the bidirectional reflectance (BR) of MLS, which fundamentally suppresses the highlight and removes the limitation on the luminous flux. To execute the proposed strategy, an unfixed view structured light system (UVSLS) is established. The UVSLS converts the reflection viewer from the real camera to the virtual camera, realizing the flexible adjustment of the MLS BR. Finally, an MLS 3D reconstruction framework is constructed to obtain the 3D profile of MLS. Experiments demonstrate that the proposed framework reduces the saturated pixels by 73.72 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> compared with the conventional method. Compared with the previous methods, the saturated pixels are reduced by an average of 47.62 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> .

Machine Learning-Based Gesture Recognition for Communication with the Deaf and Dumb

A deep learning model specifically designed to recognize signs in sign language is the foundation of the Sign Language Recognition system. Sign Language is a visual language used by the deaf and hard of hearing community to communicate with one another and the general public. Sign language is a kind of nonverbal communication based on the use of hand gestures. The ability to communicate socially and emotionally is greatly aided when the speech and hearing challenged have access to sign language. The model developed in this paper captures the images through live web cam and displays the sign language meaning on the screen as text output. The model is trained and built by deep learning framework using Convolution Neural Networks (CNN) in this work. The model is trained with images of hand gestures captured through webcam using Computer Vision and then after successful training, the system performs recognition process through matching parameters for a given input gesture and finally displays the sign language meaning of the gesture as text output on the screen.

Validation of an open source, remote web-based eye-tracking method (WebGazer) for research in early childhood.

Measuring eye movements remotely via the participant's webcam promises to be an attractive methodological addition to in-person eye-tracking in the lab. However, there is a lack of systematic research comparing remote web-based eye-tracking with in-lab eye-tracking in young children. We report a multi-lab study that compared these two measures in an anticipatory looking task with toddlers using WebGazer.js and jsPsych. Results of our remotely tested sample of 18-27-month-old toddlers (N=125) revealed that web-based eye-tracking successfully captured goal-based action predictions, although the proportion of the goal-directed anticipatory looking was lower compared to the in-lab sample (N=70). As expected, attrition rate was substantially higher in the web-based (42%) than the in-lab sample (10%). Excluding trials based on visual inspection of the match of time-locked gaze coordinates and the participant's webcam video overlayed on the stimuli was an important preprocessing step to reduce noise in the data. We discuss the use of this remote web-based method in comparison with other current methodological innovations. Our study demonstrates that remote web-based eye-tracking can be a useful tool for testing toddlers, facilitating recruitment of larger and more diverse samples; a caveat to consider is the larger drop-out rate.

Evaluation of methods for estimating autonomic nervous activity using a web camera

Evaluation of methods for estimating autonomic nervous activity using a web camera

Finding and Tracking Automobiles on Roads for Self-Driving Car Systems

Road-object detection, recognition, and tracking are vital tasks that must be performed reliably and accurately by self-driving car systems in order to achieve the automation/autonomy goal. Other vehicles are one of the main objects that the egocar must accurately detect and track on the road. However, deep-learning approaches proved their effectiveness at the expense of very demanding computational power and low throughput. They must be deployed on expensive CPUs and GPUs. Thus, in this work, a lightweight vehicle detection and tracking technique (LWVDT) is suggested to fit low-cost CPUs without sacrificing robustness, speed, or comprehension. The LWVDT is suitable for deployment in both advanced driving assistance systems (ADAS) functions and autonomous-car subsystems. The implementation is a sequence of computer-vision techniques fused together and merged with machine-learning procedures to strengthen each other and streamline execution. The algorithm details and their execution are revealed in detail. The LWVDT processes raw RGB camera pictures to generate vehicle boundary boxes and tracks them from frame to frame. The performance of the proposed pipeline is assessed using real road camera images and video recordings under different circumstances and lighting/shading conditions. Moreover, it is also tested against the well-known KITTI database, achieving an average accuracy of 87%.

Enhancing Health and Wellness: Deep Learning Powered Exercise Recognition and Tracking

Abstract: This project aims to develop a React-based web application that enables users to track, analyze, manage, and schedule their workout routines. Additionally, the application incorporates a machine learning model that utilizes the user’s web camera to track their movements and count reps, sets, and calories burned. This innovative feature provides users with a more accurate representation of their workout progress and empowers them to make data-driven decisions regarding their fitness goals. To achieve this, we utilized various technologies such as React for the frontend, Node.js for the backend, and MongoDB for the database. We also integrated TensorFlow.js, an open-source machine learning library, to build the ML model that powers the application’s tracking and analysis capabilities. Furthermore, we designed the user interface using Bootstrap with Vanilla CSS to ensure a responsive and user-friendly experience. Through this project, our primary objective was to address the limitations of existing workout tracking applications that heavily rely on subjective and often inaccurate user input. Our solution offers a more objective and efficient approach to tracking fitness progress. We firmly believe that this technology has the potential to revolutionize the fitness industry. Overall, our project showcases the power of machine learning and its ability to enhance people’s lives in meaningful ways.

High-Fidelity Drone Simulation with Depth Camera Noise and Improved Air Drag Force Models

Drone simulations offer a safe environment for collecting data and testing algorithms. However, the depth camera sensor in the simulation provides exact depth values without error, which can result in variations in algorithm behavior, especially in the case of SLAM, when transitioning to real-world environments. The aerodynamic model in the simulation also differs from reality, leading to larger errors in drag force calculations at high speeds. This disparity between simulation and real-world conditions poses challenges when attempting to transfer high-speed drone algorithms developed in the simulated environment to actual operational settings. In this paper, we propose a more realistic simulation by implementing a novel depth camera noise model and an improved aerodynamic drag force model. Through experimental validation, we demonstrate the suitability of our models for simulating real-depth cameras and air drag forces. Our depth camera noise model can replicate the values of a real depth camera sensor with a coefficient of determination (R2) value of 0.62, and our air drag force model improves accuracy by 51% compared to the Airsim simulation air drag force model in outdoor flying experiments at 10 m/s.

Determination of basic nitrogen content in diesel oil: A miniaturized method by digital image-based colorimetry in a portable device

A miniaturized method using a portable device with digital image acquisition and PhotoMetrix PRO app data treatment was developed for the determination of basic nitrogen content in diesel oil. The method was based on the colorimetric titration described in the UOP 269-10 standard protocol. A homemade 3D-printed chamber with controlled light intensity equipped with an USB camera was used for image acquisition after an acid-base titration reaction, carried out in a miniaturized device. After mixing reagents and diesel oil, the images were obtained and converted into RGB (red, green, and blue) histograms, and a partial least squares (PLS) multivariate calibration model was constructed. Parameters of the regression model were evaluated, by the coefficient of determination (R2), the root mean squared error of calibration (RMSEC), the root mean squared error of cross-validation (RMSECV), and the root mean squared error of prediction (RMSEP). Some conditions for the acid-base titration were optimized, such as the concentration of the indicator (68.0 to 272 µmol L-1) and the titrating (HClO4, 0.179 to 1.79 mmol L-1), as well as the volume of diesel oil. With 60 µL of 2.54 mmol L-1 indicator solution, 20 µL of 20 mmol L-1 HClO4 as titrating and using 50 to 1000 µL of diesel oil, optimal conditions were obtained for calibration (RMSEP of 0.377 mg kg-1, RMSECV of 0.307 mg kg-1 with 4 factors). It is important to mention that no differences were observed (p &lt; 0.05) when comparing reference values with the results by the proposed protocol. This proved to be advantageous in relation to the methods described in the UOP 269-10 standard since it was possible to reduce the consumption of reagents and waste generation, in agreement with green analytical chemistry. In addition, this alternative protocol combines simplicity and speed to obtain results with good accuracy, precision and suitable limit of quantification (1 mg kg-1) using a miniaturized system.

Didactic Tool Based on Virtual and Augmented Reality in Art Education: Learning through an Interactive Game

This article presents research on the results of using a didactic tool based on virtual and augmented reality, developed for solving tasks from the field of spatial design. The theoretical part presents the context of the tool’s development and discusses spatial perception in new media, spatial design of images and screen views, and new media and augmented reality in school practice. The didactic tool consists of a computer, a computer screen, a computer camera, and a set of markers with symbols for different architectural elements that allow different composition operations. Manipulation of the markers is captured by a computer camera, which recognizes the symbols as function keys and allows interaction with a computer program. The computer decodes the data and displays it on the screen as surfaces and solids. This strategy promotes learning like a computer game. We tested it with eighth grade elementary students. The goal was to determine the extent to which the tool would affect the implementation of artistic tasks, creativity, motivation, satisfaction, and understanding of art concepts. The results presented here are encouraging, especially in the area of creativity, which was confirmed by teachers and students after the evaluation.

How video conferencing promotes preferences for self-enhancement products

The pandemic has led to a significant increase in the use of video conferencing platforms such as Zoom, Google Meet and Microsoft Teams. Yet little is known about the impact of video conferencing on subsequent consumer decisions. Across six studies, we examine the effects of video conferencing in both consumption (e.g., sales) and nonconsumption (e.g., school and work) contexts and find that video conferencing can trigger greater interest in products that can enhance the self - physically, intellectually, and/or mentally - due to heightened social appearance anxiety. This effect is attenuated when technology allows users to reduce social appearance anxiety (e.g., the use of ring lights or the ability to turn off web cameras) but accentuated when social anxiety is increased (e.g., the use gallery/speaker views) and is more pronounced among consumers who are low in self-esteem.

Authenticity detection of eye images using deep learning technology in online exams

The use of online education technologies is one of the most effective ways to achieve the universal goal of quality education throughout life for the sustainable development of UNESCO. During the writing of the article, the features that have appread in the field of application of online exam technologies will be described. The existing forms, methods and tools for conducting the exam in an online format, their potential and risks are analyzed. Attention is drawn to the fact that the methods and tools used for conducting online control have disadvantages, such as significant costs in organizing proctoring and difficulties in assessing one of the student’s key competencies, namely, readiness to use systematized theoretical and practical knowledge to formulate and solve research problems. This research introduces an innovative proctoring system that harnesses deep learning techniques, specifically the Yolo algorithm. The primary objective of this system is to accurately detect the eyes of the exam-taker by utilizing a standard web camera as the sole input source. The obtained results from the deep learning model are subsequently subjected to processing through the Cascaded Haar Classifier to yield a definitive outcome. The proposed approach demonstrates considerable proficiency in discriminating between authentic exam-takers and static images, achieving a notable level of accuracy. Leveraging the capabilities of deep learning in conjunction with the cascaded classifier enables real-time eye detection and localization within images or video frames, thereby establishing the proctoring system as a reliable and effective solution for ensuring academic integrity during online examinations.

Image recognition software geometry with Python and OpenCV

The project aims to develop a software application using Python and various libraries for image processing and manipulation. The software used consists of Windows 10 Pro Education, Python 3.8.2, OpenCV 4.5.3, NumPy 1.22, Imutils 0.5.4, Pillow 9.4.0, Tkinter, and Visual Studio Code 1.75.1. The methodology is divided into four sprints. In the first sprint, the necessary dependencies, including OpenCV, NumPy, Imutils, Pillow, and Tkinter, were installed. The second sprint involved developing a command-line interface and integrating a live webcam video feature. In the third sprint, color detection was implemented by converting the image to black and white and applying filters to identify specific colors. Finally, the fourth sprint focused on developing the selection of geometric figures and colors. When a figure and color were selected, the software would identify and outline the corresponding object. This project utilized Python and various libraries to create an application for image processing and manipulation. The methodology followed a sprint-based approach to gradually develop different features and functionalities of the software.

Design, Implementation, and Control of a Linear Electric Actuator for Educational Mechatronics

Kinematics is a fundamental topic in engineering, robotics, mechatronics, and control systems and significantly resolves some of these fields’ most pressing issues. It is essential to assess the balance between a topic’s theoretical framework and its empirical validation to succeed in engineering. Educational tools have gained significant attention for their ability to enhance the learning experience by providing the hands-on experiences necessary to assess theoretical frameworks and empirical validations. This paper presents a system incorporating state-of-the-art features, including a fuzzy controller enabling precise control of a linear actuator and a USB camera, to provide an interactive experience. The USB camera captures the position of the actuator, providing real-time visual feedback and allowing the students to validate their theoretical understanding through practical experiments. Precision, accuracy, resolution, and the implementation of the fuzzy controller are measured to evaluate the whole system’s performance. The design, implementation, and control of our educational electrical linear actuator for teaching kinematics concepts contribute to a practical educational tool and advance interactive learning approaches in the field.