Hand-held Camera Research Articles

Clinical assessment of patients with systemic sclerosis: is there a place for thermography?

Recurrent changes of temperature and persistence of cooling along fingers at the room temperature make hands the most frequent region of interest for thermography in systemic sclerosis (SSc). The aim of this study was to evaluate dependance of temperature in hands on a subtype of the disease, immune profile of antinuclear antibodies (ANA), and lung involvement. There were 29 patients with limited cutaneous involvement (lcSSc) and 10 patients with diffuse cutaneous disease (dcSSc) enrolled for the study. To compare measurements to normal values, there were enrolled 29 healthy volunteers (control group). All participants were submitted to thermography with handheld camera FLIR One Pro for iOS, attached to mobile phone iPhone 11, at the fixed temperature of 21°C. Measurements included average temperature (Tavg) over nailfolds in thumbs and fingers II-V, as well as the difference in average temperatures (TΔ) between metacarpus of the hand and its thumb and fingers II-V. Both thumbs and fingers II-V remained cooler in subjects with dcSSc compared to those with lcSSc. This implicated a significantly greater TΔ along thumbs and fingers II-V in dcSSc group. Although Tavg at nailfolds in SSc patients was not lower than in healthy controls, TΔ remained significantly more pronounced in both lcSSc and dcSSc subjects. A positivity to ACA in lcSSc group was found to be associated with significantly lower Tavg and more pronounced TΔ in fingers II-V than the presence of anti-Scl70 antibodies. Temperature measurements remained statistically independent on a presence of ILD in lcSSc group, but both thumbs and fingers II-V in dcSSc group were warmer in case of lung involvement. The study showed the dcSSc subtype, the positivity of ACA in lcSSc, but not lung involvement were associated with poorer thermal control in the hands of SSc patients. A comparison to healthy controls highlighted the weakness of temperature measurements at nailfolds (Tavg) but increased the value of TΔ in thermography of hands.

Monte Carlo modelling of a compact CZT-based gamma camera with application to 177Lu imaging

BackgroundSemiconductor gamma-camera systems based on cadmium zinc telluride (CZT) detectors present new challenges due to an energy-response that includes effects of low-energy tailing. In particular, such energy tails produce effects that need to be considered when imaging radionuclides with multiple emissions such as ^{177}{mathrm {Lu}}. Monte Carlo simulation can be used to investigate the behaviour of such systems and optimise their use, provided that the detector model closely reflects the real physical detector. The aim of this work is to develop a CZT model applicable for simulation of CZT-based gamma cameras.MethodsThe equations describing the charge transport and signal induction are considered in three dimensions and are solved numerically, and the CZT model is then realised by coupling the detector-response to the photon-transport handled by the SIMIND Monte Carlo program. The CZT model is tuned to reproduce experimentally measured energy spectra of a hand-held gamma camera system for multiple radionuclides (^{99mathrm {m}}{mathrm {Tc}}, ^{123}{mathrm {I}} and ^{177}{mathrm {Lu}}) and parallel-hole collimators (MEGP, LEHR) as well as an uncollimated system.ResultsOverall, the model results agree well with measurements across the range of experimental conditions. The applicability of the model is demonstrated by separating energy spectra into components to investigate the interference of high-energy photons on lower energy-windows, where pronounced effects of low-energy tailing for ^{177}{mathrm {Lu}} are observed.ConclusionsThe developed model provides understanding of the specifics of the camera response and is expected to be helpful for future optimisation of gamma camera applications.

A new handheld fundus camera combined with visual artificial intelligence facilitates diabetic retinopathy screening.

To explore the performance in diabetic retinopathy (DR) screening of artificial intelligence (AI) system by evaluating the image quality of a handheld Optomed Aurora fundus camera in comparison to traditional tabletop fundus cameras and the diagnostic accuracy of DR of the two modalities. Overall, 630 eyes were included from three centers and screened by a handheld camera (Aurora, Optomed, Oulu, Finland) and a table-top camera. Image quality was graded by three masked and experienced ophthalmologists. The diagnostic accuracy of the handheld camera and AI system was evaluated in assessing DR lesions and referable DR. Under nonmydriasis status, the handheld fundus camera had better image quality in centration, clarity, and visible range (1.47, 1.48, and 1.40) than conventional tabletop cameras (1.30, 1.28, and 1.18; P<0.001). Detection of retinal hemorrhage, hard exudation, and macular edema were comparable between the two modalities, in principle, with the area under the curve of the handheld fundus camera slightly lower. The sensitivity and specificity for the detection of referable DR with the handheld camera were 82.1% (95%CI: 72.1%-92.2%) and 97.4% (95%CI: 95.4%-99.5%), respectively. The performance of AI detection of DR using the Phoebus Algorithm was satisfactory; however, Phoebus showed a high sensitivity (88.2%, 95%CI: 79.4%-97.1%) and low specificity (40.7%, 95%CI: 34.1%-47.2%) when detecting referable DR. The handheld Aurora fundus camera combined with autonomous AI system is well-suited in DR screening without mydriasis because of its high sensitivity of DR detection as well as its image quality, but its specificity needs to be improved with better modeling of the data. Use of this new system is safe and effective in the detection of referable DR in real world practice.

A new approach to non-mydriatic portable fundus imaging

ABSTRACT Introduction The present study proposes a new hand-held non-mydriatic fundus camera for retinal imaging. The goal is to design a fundus camera which is equally effective in both clinical and telemedicine scenarios. Areas covered A new retinal illumination approach is proposed to address the main dilemma of the optical design, i.e. balancing efficacy with structural simplicity. This is achieved by symmetrical and co-axial placement of multiple illumination sources along the optical pathway. Each illumination source includes a white and a Near Infra-Red (NIR) LED, which are placed adjacent to each other. Hence, the camera can produce a view-finder with NIR illumination without the need for additional beam-splitters and filters. Expert Opinion The proposed design blends the structural simplicity of the ‘off-axis illumination with the wide field of view and uniform illumination of the ‘ring’ illumination. Moreover, the camera is designed to work with Android-based smartphones, which can easily be mounted and interfaced. The efficacy of the proposed camera is determined by ocular safety analysis and comparative evaluation with a table-top fundus camera. The results convincingly demonstrate the ability of the proposed camera as a primary driver of a wide-scale screening program in both clinical and remote resource constraint environments.

Assessment of peach trees water status and leaf gas exchange using on-the-ground versus airborne-based thermal imagery

The thermal region of the electromagnetic spectrum might provide valuable information for assessing plant water status. Nevertheless, the plant’s physiology and the scale of measurement, (e.g. sensor viewing geometry and the canopy aggregation) are critical for quantifiying and monitoring water stress. This study compares the Crop Water Stress Index (CWSI) of a peach orchard obtained using on-the-ground, and airborne-based canopy temperature (Tc). The temporal evolution of CWSI under mild water stress conditions was assessed for three different irrigation strategies (over-irrigation, OI; farmer irrigation, FI; and non-irrigation, NI). Two aerial campaigns per irrigation season (2017–2018) were performed with an airborne thermal sensor: a first flight under well-watered conditions, and a second flight once mild water stress was developed. At the time of the flights, Tc and net photosynthesis (Pn), stomatal conductance (gs) and stem water potential (Ψs) were measured on the ground with a hand-held thermal camera, a portable gas exchange system and a pressure chamber, respectively. The canopy temperature obtained from the hand-held thermal camera, averaging the sunlit and shaded parts of the canopy, agreed with that derived from the airborne measurements (Y=1.00X; RMSE= 1.97 K). The CWSI values calculated from both approaches detected peach water status under different irrigation strategies. In general, Ψs was better predicted from the aircraft (R² up to 0.72 for CWSI obtained from the aircraft versus R2 =0.51 for Tc ground measurements), whereas the use of ground measurements was preferred for estimating gs and Pn (R² up to 0.73 and 0.74 for Tc ground measurements versus R2 =45 and 0.56 for Tc and CWSI derived from the aircraft). Regardless the approach used for deriving Tc, and due to the consideration of different meteorological conditions (i.e different dates), CWSI provided a better relationship with Ψs than Tc, whereas the latter was more closely related with gs and Pn.

Move: Durational Dance-Making as We Age.

Dance making and moving support a positive aging approach to aesthetic, cultural, and social meanings, thus, shaping dancers' lives. My perspective is informed by making dance a durational and sustained practice in the life of a long-life mover. Positioned as a practice-oriented researcher, I am surrounded by inspiring authors and artists who embrace maturity with improvisational making. This perspective piece reflects on making Move, a durational practice-oriented research process. Move contextualizes durational elements that I associates with the life of a mover; my writing and my companions improvising with my hand-held camera in isolation, in my neighborhood on Tāmaki Makaurau/Auckland's North Shore.This perspective piece opens a process I refer to as a camera-dancer dyad, a duet between a dancer and a camera, making in isolation as we age. Furthermore, by embracing the accessibility of everyday recording devices as dancing partners, dance making with mobile cameras becomes a playful partnership with a long life.

Biological Tomato Leaf Disease Classification using Deep Learning Framework

Biological Tomato leaf classification is very important to decide the pesticide, insecticide, and other treatments needed for the plant to yield good crop. The images captured by handheld cameras or using drones are used by various machine learning algorithms to identify the diseases. Such methods need extraction of features from the images before the machine learning methods can be used for disease identification. In this paper, a deep learning framework is proposed that automatically extracts features in a hierarchical manner. The features are classified using neural networks to classify the leaves into three classes, viz. no disease, bacterial spot, and Septoria leaf spot. The performance of the model is tested using accuracy as the performance metric. The obtained performance metric validates the performance of the method. The method is useful for taking corrective measures to disease management of tomato plants.

A Novel Online Dissection Course on Lower Limb Anatomy During the COVID-19 Pandemic.

Introduction: The teaching of human anatomy, a medical subject that relies heavily on live teaching, teacher-student interactivity, and visuospatial skills, has suffered tremendously since the COVID-19 pandemic mandated the shutting down of medical institutions. The medical education fraternity was compelled to replace the traditional teaching method of hands-on cadaveric dissections (HOCDs) with online education to overcome this new challenge, but it came at the cost of reduced student engagement and lesser spatial orientation.Method: In this cross-sectional, questionnaire-based study, we designed a novel online dissection course on lower limb anatomy and collected student feedback on the same from consenting Phase I Bachelor of Medicine, Bachelor of Surgery (MBBS) students of Symbiosis Medical College for Women, Pune, India. The course design consisted of three different modes: a live Zoom session using a handheld camera phone, a pre-recorded video dissection uploaded on the institute learning management system, and a Powerpoint presentation with high-resolution photographs of each dissected layer; and the feedback intended to find out what works best for the students. Overall feedback regarding their preferences in terms of presentation design, use of background music in pre-recorded videos, and overall learning experience was also collected. The course consisted of six two-hour teaching sessions. The first three sessions each used a different mode of teaching, repeating the same pattern in the next three sessions. The first mode of teaching implemented was a live Zoom session where instructors used a hand-held cell phone camera to show specimens that had been dissected a day prior. The second mode involved a pre-recorded video showing step-by-step dissection performed by the instructor which was then uploaded on the Institute Learning Management System. Of the two pre-recorded videos, background music consisting of a low-volume instrumental track was added to the second video. The third mode utilized Powerpoint presentations containing high-resolution photographs of each dissected layer on a separate slide along with labeling. The presentations were shown to the students over a Zoom call. A Google Form (GF) questionnaire was created after validation by subject experts to gather the students’ feedback on the teaching and learning of anatomy via these sessions. The GF responses were collected and analyzed using Microsoft Excel.Results: 41.7% of students recommended the use of a combination of all three modes in the same session, while 36.7% favored pre-recorded videos. 86.7% of students said that a good quality presentation design helps in keeping them engaged and only 23% of students favored the use of background music for increasing their ability to concentrate. 63.3% of students found the learning experience highly satisfactory.Conclusion: Although virtual dissection teaching methods may not be able to completely replace HOCDs, a well-planned online dissection course incorporating multiple modes of online dissections with an emphasis on good quality presentation design and frequent teacher-student interactivity can provide a strong impetus for learning in the absence of live teaching methods.

Comparison of Handheld Retinal Imaging with ETDRS 7-Standard Field Photography for Diabetic Retinopathy and Diabetic Macular Edema

To compare nonmydriatic (NM) and mydriatic (MD) handheld retinal imaging with standard ETDRS 7-field color fundus photography (ETDRS photographs) for the assessment of diabetic retinopathy (DR) and diabetic macular edema (DME). Prospective, comparative, instrument validation study. A total of 225 eyes from 116 patients with diabetes mellitus. Following a standardized protocol, NM and MD images were acquired using handheld retinal cameras (NM images: Aurora, Smartscope, and RetinaVue-700; MD images: Aurora, Smartscope, RetinaVue-700, and iNview) and dilated ETDRS photographs. Grading was performed at a centralized reading center using the International Clinical Classification for DR and DME. Kappa statistics (simple [K], weighted [Kw]) assessed the level of agreement for DR and DME. Sensitivity and specificity were calculated for any DR, referable DR (refDR), and vision-threatening DR (vtDR). Agreement for DR and DME; sensitivity and specificity for any DR, refDR, and vtDR; ungradable rates. Severity by ETDRS photographs: no DR, 33.3%; mild nonproliferative DR, 20.4%; moderate DR, 14.2%; severe DR, 11.6%; proliferative DR, 20.4%; no DME, 68.0%; DME, 9.3%; non-center involving clinically significant DME, 4.9%; center-involving clinically significant DME, 12.4%; and ungradable, 5.3%. For NM handheld retinal imaging, Kw was 0.70 to 0.73 for DR and 0.76 to 0.83 for DME. For MD handheld retinal imaging, Kw was 0.68 to 0.75 for DR and 0.77 to 0.91 for DME. Thresholds for sensitivity (0.80) and specificity (0.95) were met by NM images acquired using Smartscope and MD images acquired using Aurora and RetinaVue-700 cameras for any DR and by MD images acquired using Aurora and RetinaVue-700 cameras for refDR. Thresholds for sensitivity and specificity were met by MD images acquired using Aurora and RetinaVue-700 for DME. Nonmydriatic and MD ungradable rates for DR were 15.1% to 38.3% and 0% to 33.8%, respectively. Following standardized protocols, NM and MD handheld retinal imaging devices have substantial agreement levels for DR and DME. With mydriasis, not all handheld retinal imaging devices meet standards for sensitivity and specificity in identifying any DR and refDR. None of the handheld devices met the established 95% specificity for vtDR, suggesting that lower referral thresholds should be used if handheld devices must be utilized. When using handheld devices, the ungradable rate is significantly reduced with mydriasis and DME sensitivity thresholds are only achieved following dilation.

Abstract 147: The diagnostic accuracy of portable, non-contact, handheld retinal camera in screening diabetic retinopathy

Background: Hand-held retinal camera (HRC) is a portable device which require less expertise and training with less image acquisition time in identifying diabetic retinopathy (DR).Aims and Objectives: To find out the diagnostic accuracy of HRC in identifying DR in comparison to the table top fundus camera (TTFC).Materials and Methods: Adults with diabetes mellitus irrespective of the type and duration, attending the ophthalmology clinic were randomly selected and initially subjected to retinal photography with TTFC on a dilated eye individually by an expert ophthalmologist followed by retinal photography with HRC by a trained resident. The best images of both the devices were graded by ophthalmologist according to the ETDRS classification with proper blinding of the photographs.Results: 101 subjects (191 eyes) with mean age of 56.86 + 10.75 years were evaluated. 45 eyes with Non Proliferative DR (NPDR) and 1 eye with Proliferative DR (PDR) were diagnosed with TTFC among which HRC identified retinopathy in 34 eyes (33 NPDR & 1 PDR). All 19 cases of Diabetic Macular edema (DME) were identified by HRC. HRC had sensitivity and specificity of 79.9% and 99.3% in diagnosing DR with a positive predictive value of 97.1%.Conclusion: HRC using mydriasis has good diagnostic accuracy in preliminary DR screening.

RectiNet-v2: A stacked network architecture for document image dewarping

• A Gated and Bifurcated stacked network is proposed to rectify warped documents. • Residual Paths are introduced to enhance the flow of information within the network. • A novel boundary weighted loss is used to enable greater focus on boundaries. • Results indicating dewarp quality are summarized with relevant comparisons. With the advent of mobile and hand-held cameras, document images have found their way into almost every domain. Dewarping of these images for the removal of perspective distortions and folds is essential so that they can be understood by document recognition algorithms . For this, we propose an end-to-end CNN architecture that can produce distortion-free document images from warped documents it takes as input. We train this model on warped document images simulated synthetically to compensate for the lack of enough natural data. Our method is novel in the use of a bifurcated decoder with shared weights to prevent intermingling of grid coordinates, in the use of residual networks in the U-Net skip connections to allow the flow of data from different receptive fields in the model, and in the use of a gated network to help the model focus on structure and line-level detail of the document image. We evaluate our method on the DocUNet dataset, a benchmark in this domain, and obtain results comparable to state-of-the-art methods.

Visual–inertial structural acceleration measurement

AbstractStructural vibration measurement is a crucial and necessary step for structural health monitoring. Recently, computer vision‐based techniques have been proposed by researchers to measure structural motion remotely. However, the direct application of vision‐based measurement to practical applications still faces some challenges, mainly because intrinsic camera vibration can introduce significant errors to the measurement results. In this study, a three‐stage approach using an embedded inertial measurement unit is proposed to compensate for the camera movement. First, camera rotations are estimated by employing a complementary filter with an adaptive gain to fuse gyroscope measurement and accelerometer data. Next, binary robust invariant scalable key‐point features are detected from the region of interest and tracked between video frames using a Kanade–Lucas–Tomasi tracker. Finally, structural acceleration is obtained by combining the information for the obtained structural features and the estimated nonstationary camera motion. The performance of the proposed approach is investigated using both a moving handheld camera and a camera mounted on the unmanned aerial vehicle in the laboratory. These results demonstrate that the proposed method can be effectively applied to measure structural vibration, without requiring stationary background features to be available in the video.

PHOTOGRAMMETRY, LASER SCANNING AND HBIM FOR CONSTRUCTION DIAGNOSTIC

Abstract. The aim of this article is to show today's possibilities of the documentation of damaged monuments by various technologies on a case study. Laser scanning, mobile laser scanning, terrestrial photogrammetry using hand-held camera, and drone aerial photogrammetry were used. The outputs were textured and non-textured point clouds and an aerial orthophoto of the object. The clouds were compared in terms of possible deformations, the issues of overall accuracy and time required to focus the object were also discussed. The outputs were used for the basic plans of the building, intended for the planned reconstruction and also for the creation of a model for HBIM.

Image-Based Differentiation of Bacterial and Fungal Keratitis Using Deep Convolutional Neural Networks.

PurposeDevelop computer vision models for image-based differentiation of bacterial and fungal corneal ulcers and compare their performance against human experts.DesignCross-sectional comparison of diagnostic performance.ParticipantsPatients with acute, culture-proven bacterial or fungal keratitis from 4 centers in South India.MethodsFive convolutional neural networks (CNNs) were trained using images from handheld cameras collected from patients with culture-proven corneal ulcers in South India recruited as part of clinical trials conducted between 2006 and 2015. Their performance was evaluated on 2 hold-out test sets (1 single center and 1 multicenter) from South India. Twelve local expert cornea specialists performed remote interpretation of the images in the multicenter test set to enable direct comparison against CNN performance.Main Outcome MeasuresArea under the receiver operating characteristic curve (AUC) individually and for each group collectively (i.e., CNN ensemble and human ensemble).ResultsThe best-performing CNN architecture was MobileNet, which attained an AUC of 0.86 on the single-center test set (other CNNs range, 0.68–0.84) and 0.83 on the multicenter test set (other CNNs range, 0.75–0.83). Expert human AUCs on the multicenter test set ranged from 0.42 to 0.79. The CNN ensemble achieved a statistically significantly higher AUC (0.84) than the human ensemble (0.76; P < 0.01). CNNs showed relatively higher accuracy for fungal (81%) versus bacterial (75%) ulcers, whereas humans showed relatively higher accuracy for bacterial (88%) versus fungal (56%) ulcers. An ensemble of the best-performing CNN and best-performing human achieved the highest AUC of 0.87, although this was not statistically significantly higher than the best CNN (0.83; P = 0.17) or best human (0.79; P = 0.09).ConclusionsComputer vision models achieved superhuman performance in identifying the underlying infectious cause of corneal ulcers compared with cornea specialists. The best-performing model, MobileNet, attained an AUC of 0.83 to 0.86 without any additional clinical or historical information. These findings suggest the potential for future implementation of these models to enable earlier directed antimicrobial therapy in the management of infectious keratitis, which may improve visual outcomes. Additional studies are ongoing to incorporate clinical history and expert opinion into predictive models.

Evaluating a Deep Learning Diabetic Retinopathy Grading System Developed on Mydriatic Retinal Images When Applied to Non-Mydriatic Community Screening.

Artificial Intelligence has showcased clear capabilities to automatically grade diabetic retinopathy (DR) on mydriatic retinal images captured by clinical experts on fixed table-top retinal cameras within hospital settings. However, in many low- and middle-income countries, screening for DR revolves around minimally trained field workers using handheld non-mydriatic cameras in community settings. This prospective study evaluated the diagnostic accuracy of a deep learning algorithm developed using mydriatic retinal images by the Singapore Eye Research Institute, commercially available as Zeiss VISUHEALTH-AI DR, on images captured by field workers on a Zeiss Visuscout® 100 non-mydriatic handheld camera from people with diabetes in a house-to-house cross-sectional study across 20 regions in India. A total of 20,489 patient eyes from 11,199 patients were used to evaluate algorithm performance in identifying referable DR, non-referable DR, and gradability. For each category, the algorithm achieved precision values of 29.60 (95% CI 27.40, 31.88), 92.56 (92.13, 92.97), and 58.58 (56.97, 60.19), recall values of 62.69 (59.17, 66.12), 85.65 (85.11, 86.18), and 65.06 (63.40, 66.69), and F-score values of 40.22 (38.25, 42.21), 88.97 (88.62, 89.31), and 61.65 (60.50, 62.80), respectively. Model performance reached 91.22 (90.79, 91.64) sensitivity and 65.06 (63.40, 66.69) specificity at detecting gradability and 72.08 (70.68, 73.46) sensitivity and 85.65 (85.11, 86.18) specificity for the detection of all referable eyes. Algorithm accuracy is dependent on the quality of acquired retinal images, and this is a major limiting step for its global implementation in community non-mydriatic DR screening using handheld cameras. This study highlights the need to develop and train deep learning-based screening tools in such conditions before implementation.

All-In-Focus Polarimetric Imaging Based on an Integrated Plenoptic Camera with a Key Electrically Tunable LC Device.

In this paper, a prototyped plenoptic camera based on a key electrically tunable liquid-crystal (LC) device for all-in-focus polarimetric imaging is proposed. By using computer numerical control machining and 3D printing, the proposed imaging architecture can be integrated into a hand-held prototyped plenoptic camera so as to greatly improve the applicability for outdoor imaging measurements. Compared with previous square-period liquid-crystal microlens arrays (LCMLA), the utilized hexagonal-period LCMLA has remarkably increased the light utilization rate by ~15%. Experiments demonstrate that the proposed imaging approach can simultaneously realize both the plenoptic and polarimetric imaging without any macroscopic moving parts. With the depth-based rendering method, both the all-in-focus images and the all-in-focus degree of linear polarization (DoLP) images can be obtained efficiently. Due to the large depth-of-field advantage of plenoptic cameras, the proposed camera enables polarimetric imaging in a larger depth range than conventional 2D polarimetric cameras. Currently, the raw light field images with three polarization states including I0 and I60 and I120 can be captured by the proposed imaging architecture, with a switching time of several tens of milliseconds. Some local patterns which are selected as interested target features can be effectively suppressed or obviously enhanced by switching the polarization state mentioned. According to experiments, the visibility in scattering medium can also be apparently improved. It can be expected that the proposed polarimetric imaging approach will exhibit an excellent development potential.

Smart building demolition and waste management frame with image-to-BIM

Increasing urbanization has resulted in a large number of building demolition, generating massive volumes of construction and demolition waste (CDW) in many countries. However, adjustments and deterioration to existing structures are poorly documented, yielding a data gap before the demolition work. Additionally, the entire demolition process relies heavily on on-site operations, pre-demolition audit and efficient construction and demolition waste management (CDWM). To improve the efficiency and informatization of building demolition and CDWM, this article proposes a smart framework based on Image-to-BIM technique. Within this frame, a combined UAV and handheld camera system is presented to collect interior and exterior images of to-be-demolished buildings automatically and completely, in which camera motion path planning is accomplished by considering the target region. The proposed algorithm could detect the line features in the image-based reconstruction model and extracts geometric data to get the required BIM model. Additional frame components dedicated to demolition planning and waste management enable schedule optimizing, safety assessment, CDW quantity take-off, and economic estimation using the data from the BIM model. The application of the framework in a masonry structure showcases how the integrated system improves the information flow during building deconstruction and CDWM, realizing a potentially smart deconstruction activity.

Application of Image Mosaic Technology in Tai Chi Animation Creation.

Panorama can reflect the image seen at any angle of view at a certain point of view. How to improve the quality of panorama stitching and use it as a data foundation in the “smart tourism” system has become a research hotspot in recent years. Image stitching means to use the overlapping area between the images to be stitched for registration and fusion to generate a new image with a wider viewing angle. This article takes the production of “Tai Chi” animation as an example to apply image stitching technology to the production of realistic 3D model textures to simplify the production of animation textures. A handheld camera is used to collect images in a certain overlapping area. After cylindrical projection, the Harris algorithm based on scale space is adopted to detect image feature points, the two-way normalized cross-correlation algorithm matches the feature points, and the algorithm to extract the threshold T iteratively removes mismatches. The transformation parameter model is quickly estimated through the improved RANSAC algorithm, and the spliced image is projected and transformed. The Szeliski grayscale fusion method directly calculates the grayscale average of the matching points to fuse the image, and finally, the best stitching method is used to eliminate the ghosting at the image mosaic. Data experiments based on Matlab show that the proposed image splicing technology has the advantages of high efficiency and clear spliced images and a more satisfactory panoramic image visual effect can be achieved.

Video-Based Pose Estimation for Gait Analysis in Stroke Survivors during Clinical Assessments: A Proof-of-Concept Study

Recent advancements in deep learning have produced significant progress in markerless human pose estimation, making it possible to estimate human kinematics from single camera videos without the need for reflective markers and specialized labs equipped with motion capture systems. Such algorithms have the potential to enable the quantification of clinical metrics from videos recorded with a handheld camera. Here we used DeepLabCut, an open-source framework for markerless pose estimation, to fine-tune a deep network to track 5 body keypoints (hip, knee, ankle, heel, and toe) in 82 below-waist videos of 8 patients with stroke performing overground walking during clinical assessments. We trained the pose estimation model by labeling the keypoints in 2 frames per video and then trained a convolutional neural network to estimate 5 clinically relevant gait parameters (cadence, double support time, swing time, stance time, and walking speed) from the trajectory of these keypoints. These results were then compared to those obtained from a clinical system for gait analysis (GAITRite®, CIR Systems). Absolute accuracy (mean error) and precision (standard deviation of error) for swing, stance, and double support time were within 0.04 ± 0.11 s; Pearson’s correlation with the reference system was moderate for swing times (r = 0.4–0.66), but stronger for stance and double support time (r = 0.93–0.95). Cadence mean error was −0.25 steps/min ± 3.9 steps/min (r = 0.97), while walking speed mean error was −0.02 ± 0.11 m/s (r = 0.92). These preliminary results suggest that single camera videos and pose estimation models based on deep networks could be used to quantify clinically relevant gait metrics in individuals poststroke, even while using assistive devices in uncontrolled environments. Such development opens the door to applications for gait analysis both inside and outside of clinical settings, without the need of sophisticated equipment.

Fruits and Vegetables Freshness Categorization Using Deep Learning

The nutritional value of perishable food items, such as fruits and vegetables, depends on their freshness levels. The existing approaches solve a binary class problem by classifying a known fruit\vegetable class into fresh or rotten only. We propose an automated fruits and vegetables categorization approach that first recognizes the class of object in an image and then categorizes that fruit or vegetable into one of the three categories: pure-fresh, medium-fresh, and rotten. We gathered a dataset comprising of 60K images of 11 fruits and vegetables, each is further divided into three categories of freshness, using hand-held cameras. The recognition and categorization of fruits and vegetables are performed through two deep learning models: Visual Geometry Group (VGG-16) and You Only Look Once (YOLO), and their results are compared. VGG-16 classifies fruits and vegetables and categorizes their freshness, while YOLO also localizes them within the image. Furthermore, we have developed an android based application that takes the image of the fruit or vegetable as input and returns its class label and its freshness degree. A comprehensive experimental evaluation of proposed approach demonstrates that the proposed approach can achieve a high accuracy and F1score on gathered <i>FruitVeg Freshness</i> dataset. The dataset is publicly available for further evaluation by the research community.