Reference Photos Research Articles

Wireless image transfer by various antennas using transmitter and receiver modules at 5.8 GHz

Abstract This research proposes an inexpensive technique for wireless image transfer for security and surveillance applications. The technique uses a 5.8 GHz transmitter and receiver module, along with external antennas in the real-time image transfer within a radius of 100 m. The transferred images are stored in a laptop using a Python code-based graphical user interface application. Different antennas, dipole, circular split-ring resonators, hexagonal split-ring resonators, and metamaterial antennas are utilized for comparison. The Blind/Referenceless Image Spatial Quality Evaluator method is used to assess the picture quality of transferred images to quantify image transfer performance when no ground truth or reference photos are supplied. According to the presented results, images transferred using metamaterial antennas have higher quality than those transferred with other types of antennas. For security considerations, such a system can communicate and store the images in real time.

Impacts of necrotising disease on the Endangered cauliflower soft coral (Dendronephthya australis)

Context Diseases have affected coral populations worldwide, leading to population declines and requiring active restoration efforts. Aims Describe population and individual impacts of necrotising disease in the Endangered octocoral Dendronephthya australis. Methods We quantified population loss and recruitment by using reference photos, survey and GPS mapping and described disease lesions by using histopathology. Key results From December 2019 to January 2020, we observed polyp loss, necrotic lesions and loss of large colonies of D. australis at Botany Bay, New South Wales, Australia. By September 2020, only a few scattered recruits remained, and all large colonies were lost. Histopathology of colonies sampled in January 2020 confirmed that the disease had resulted in necrosis, gastrovascular canal collapse and internal colony integrity loss, leading to mortality. New recruits were recorded within 10 months of disease onset, and large colonies within 18 months. Conclusions Although the necrotising disease had significant impacts on both the individual and population level, natural recruitment began quickly. As such, unlike in other populations, restoration is not currently required in the Bare Island D. australis population. Implications The extent of disease impact at the individual and population levels suggests that monitoring for lesions should be undertaken before developing conservation and restoration strategies for this species.

Snap Decisions: Assessing Participation and Data Quality in a Citizen Science Program Using Repeat Photography

Photo-point monitoring through repeat photography allows assessment of long-term ecosystem changes, and photos may be collected using citizen science methods. Such efforts can generate large photo collections, but are susceptible to varying participation and data quality. To date, there have been few assessments of the success of citizen science projects using repeat photography methods in meeting their objectives. We report on the success of the PhotoMon Project, a photo-point monitoring program at Pinery Provincial Park, Canada, at meeting its primary goals of affordably collecting seasonal reference photographs of significant ecosystems within the park, while providing a stewardship opportunity for park visitors. We investigated how the quantity of submitted photos varied over time (quantity), and how closely those photos matched the suite of criteria of the PhotoMon Project (quality). Photo submissions occurred year-round and at all sites, although a low proportion of park visitors participated in the program. Photo quantity varied among sites and seasonally, reaching a low during the winter, but with proportional participation in the project lowest in summer. Photo quality was consistent year-round, with most photos meeting most program criteria. Common issues with photo quality included photo lighting and orientation. We conclude that the program met its scientific goal of compiling seasonal reference photos, but that comparatively few park visitors engage in the program. We suggest changes to increase visitor motivation to participate, but recognize that these may compromise the program’s current affordability and ease of management.

Combining Camera Trap Data and Environmental Data to Estimate the Effects of Environmental Gradients on Abundance of the Asian Elephant Elephas maximus in Cambodia

Asian elephant (Elephas maximus) populations in Cambodia are currently declining, and the effect of environmental degradation on the abundance and health of elephants is poorly understood. We used camera trap data from 42 locations between 2016 to 2020 in the southern Cardamom Mountains to investigate the impact of environmental degradation on the abundance and condition of Asian elephants. Camera trap data were organized using CameraSweet software to retrieve both number of individuals and their condition. For a number of individuals, we defined independent captures spatially and temporally. To assess condition, we created a visual scoring system based on past research (Wemmer et al. 2006, Fernando et al. 2009, Morfeld et al. 2014, Wijeyamohan et al. 2014, Morfeld et al. 2016, Schiffmann et al. 2020). This scoring system relies on visual assessment of the muscle and fat in relation to the pelvis, ribs, and back bone. To validate this subjective scoring system, two scorers reviewed elephant captures by using 10 reference photos and then reviewing each other’s assessment in the first five images showing the elephant's body condition. This method minimizes subjective assessment from two scorers. Environmental variables (Suppl. material 1) such as distance to forest edge, forest integrity index, elevation, global human settlements, distance to road, distance to river, night light and forest cover were obtained, then reclassified in ArcGIS to a common 1 km grid. We implemented hierarchical N-mixture models to investigate the impacts of environmental variables on abundance and used cumulative link models to investigate the impact of the same environmental variables on condition. We found that Asian elephant abundance exhibited a significant positive relationship with distance to forest edges, where abundance was greater further away from a forest edge. We found that body condition score exhibited the relationship with forest cover and Forest Landscape Integrity Index, which suggested that grassland and less dense forest support better condition. Moreover, males exhibited significantly higher scores for body condition than females, while babies, juveniles, and subadults all exhibited lower body condition scores compared to adults. The significantly lower body condition of young elephants is concerning and suggests that conservation managers in the region should prioritize environmental conditions that support young elephant health. Our results identify key environmental variables that appear to promote Asian elephant abundance and health in the Cardamom Mountains, thus informing relevant conservation actions to support this endangered species in Cambodia and beyond.

A semi‐automated camera trap distance sampling approach for population density estimation

AbstractCamera traps have become important tools for the monitoring of animal populations. However, the study‐specific estimation of animal detection probabilities is key if unbiased abundance estimates of unmarked species are to be obtained. Since this process can be very time‐consuming, we developed the first semi‐automated workflow for animals of any size and shape to estimate detection probabilities and population densities. In order to obtain observation distances, a deep learning algorithm is used to create relative depth images that are calibrated with a small set of reference photos for each location, with distances then extracted for animals automatically detected by MegaDetector 4.0. Animal detection by MegaDetector was generally independent of the distance to the camera trap for 10 animal species at two different study sites. If an animal was detected both manually and automatically, the difference in the distance estimates was often minimal at a distance about 4 m from the camera trap. The difference increased approximately linearly for larger distances. Nonetheless, population density estimates based on manual and semi‐automated camera trap distance sampling workflows did not differ significantly. Our results show that a readily available software for semi‐automated distance estimation can reliably be used within a camera trap distance sampling workflow, reducing the time required for data processing, by >13‐fold. This greatly improves the accessibility of camera trap distance sampling for wildlife research and management.

The Study of a Validated Assessment Scale for a Localized Submental Fat Volume.

This aim of this study was to develop an objective tool for rating submental fat applied to Koreans. The study was conducted between April 2019 and October 2019. A total of 92 subjects were enrolled in the study. Clinical photos of the subjects were categorized using validated CR-SMFRS by three plastic surgeons and one dermatologist. The categorized photos were then shown to six different plastic surgeons for evaluation. The Cohen's kappa value for the six raters were 0.830, 0.742, 0.703, 0.907, 0.862, and 0.793 with statistical significance (p < 0.001). ICC value was between 0.860 and 0.966 (p < 0.001). Since the Cohen's value and ICC were above 0.6 for all raters, the ratings performed by all six raters were used in the analysis. The ICC values between raters were between 0.899 and 0.902. We came up with a set of reference photos that can be used for submental fat rating scale applicable to Korean subjects. II.

Разработка классификатора фотоизображений патологий для ультрамалого набора данных

The purpose of the work: to create an algorithm and implement it in a software tool for classifying photographic images of pathology of the central region of the human fundus, detected by autofluorescence research, according to 8 types-patterns: normal, minimal changes, focal, spotted, linear, lace-like, reticular, speckled. Methods: machine learning algorithms (convolutional neural networks) and computer vision (histogram methods, perceptual hash algorithms). The main feature of the task: an ultra-small set of unique photoimages with an accurately diagnosed type of pathology (18 pieces). The accuracy of forecasts when solving a problem using a neural network is 12.5%. The accuracy of the predictions of the developed algorithm using a combination of histograms, perceptual hash and 1 reference photo of the normal state of the fundus is 60% when selecting the classifier parameters from a set of 1 photo for 1 pathology. When using 3 reference photos, the norm is 85%. The proposed solution can be used in medicine, ophthalmology, photonics and optics of biological tissues, machine learning for both research and educational purposes.

Dr Salman visiting his family

When Dr Salman sent me some reference photos, I instantly knew I had to choose the photo this painting is based on - although it was a complex scene with several figures, the narrative it presented was very powerful and emotional. I wanted to avoid sentimentality, and I think that the bold impasto technique I used, with thick oil paint and a large brush, helped with this. It was mostly about conveying raw emotion in an expressive way, without including lots of detail. Also making it an easily readable image just using loose brushstrokes, without delineating everything was a real challenge. I hope I succeeded.&#x0D;

A Utility Model for Photo Selection in Mobile Crowdsensing

Existing mobile photo crowdsensing approaches focus on the participant-to-server photo pre-selection, i.e., reducing the photo redundancy from participants to a server. The server may still receive plenty of photos for a target area. Yet, another important problem is to select a proper photo subset of an area from the server to a requester. This is a challenging problem because the selected subset with a small size should attain both coverage on the PoIs - Points of Interest (i.e., photo coverage of the area) and quality on the views (i.e., view quality). In this paper, we propose a novel and generic server-to-requester photo selection approach even when there are neither photo shooting direction information nor reference photos. A utility model is designed to measure photo merits of coverage and quality by exploiting photos' spatial distribution and visual representativeness. We present two photo selection schemes, basic and PoI number-aware, to maximize the photo selection utility with multiple levels of granularity. Experimental results on real-world datasets show that our basic scheme outperforms the baselines by an average of 33% and 18.7% on photo coverage and view quality, respectively. Our PoI number-aware scheme can yield an additionally 44.8 percent improvement on the photo coverage performance.

Parental reference photos do not always improve the accuracy of forensic age progressions

During long-term missing children cases, forensic artists construct age-progressions to estimate the child’s current appearance. It is commonly believed that incorporating information about the child’s biological relatives is critical in accurately estimating the child’s current appearance. However, some evidence suggests that predicting appearance based on inheritance of features may be error prone. The present studies examine whether age-progressions constructed with the aid of a biological reference photos led to better recognition than those constructed without a biological reference. We also investigated whether there would be any variation depending on the age-range of the age-progressions. Eight professional forensic artists created age-progressions based upon photographs provided by each of our eight targets. Half of their age progressions with the aid of parental reference photos and half without parental reference photos. Furthermore, half were age-progressed across a longer age-range (5–20 years) and half covered a shorter age-range (12–20 years). In Experiment 1 similarity scores were higher over shorter age-ranges. Further, across longer age-ranges age-progressions created with the aid of a parental reference were lower than those without a reference. In Experiment 2 recognition performance was higher across shorter age-ranges. Additionally, across longer age-ranges age-progressions created with the aid of a parental reference were recognized worse than those without a reference. These results suggest that in long-term missing person cases, forensic artists may benefit from not relying on biological references. Finally, consistent with previous research (e.g. Lampinen et al., 2012) age-progressions provided no benefit over using outdated photographs.

MichiGAN

Despite the recent success of face image generation with GANs, conditional hair editing remains challenging due to the under-explored complexity of its geometry and appearance. In this paper, we present MichiGAN (Multi-Input-Conditioned Hair Image GAN), a novel conditional image generation method for interactive portrait hair manipulation. To provide user control over every major hair visual factor, we explicitly disentangle hair into four orthogonal attributes, including shape, structure, appearance, and background. For each of them, we design a corresponding condition module to represent, process, and convert user inputs, and modulate the image generation pipeline in ways that respect the natures of different visual attributes. All these condition modules are integrated with the backbone generator to form the final end-to-end network, which allows fully-conditioned hair generation from multiple user inputs. Upon it, we also build an interactive portrait hair editing system that enables straightforward manipulation of hair by projecting intuitive and high-level user inputs such as painted masks, guiding strokes, or reference photos to well-defined condition representations. Through extensive experiments and evaluations, we demonstrate the superiority of our method regarding both result quality and user controllability.

"Halmoni" and "Paternal Grandmother"

"Halmoni" and "Paternal Grandmother" Monica Heilman (bio) ARTIST STATEMENT. In this set of portraits, I depict my grandmothers, both living, one residing in South Korea, the other in small town USA. During the pandemic, my mind is frequently on elder family members. These portraits honor and memorialize, while capturing our current reality, when one's likeness is most likely obscured by a mask. Yet the mask is contentious. I am confident that my halmoni must wear one, and almost as confident that my conservative white grandmother does not. In recreating the image of a woman who does not want to see me, I contend with the tensions of a multiracial Asian American identity for which long-standing racism is inseparable from ancestry. [End Page 319] Click for larger view View full resolution Figure 1. Halmoni Deciding on your portrait is easy. Rough pencil strokes that act as pathways for careful ink lines are external expressions of longing. I reference photos I've taken of us, but these are already four years old. So I lean on a more recent photo of you and the daughter you're having difficulty remembering now—my mother—taking note of how the nursing home has dressed you in something akin to a hospital gown. I know that now you must also wear a mask, because you live in a country that acknowledges that public health takes communal effort. I take comfort in that fact, drawing in the shadows of your wrinkles as I wonder if I'll get to see you again. [End Page 320] Click for larger view View full resolution Figure 2. Paternal Grandmother As soon as I decided to do your portrait, I questioned whether I could pull it off. We haven't seen each other in twenty years and, currently quarantined away from my childhood home, I have no photos of you. You hate photos anyway. While Halmoni's portrait leaves warmth, drawing yours highlights an emptiness. I pull on faded memories of your likeness and do a Google image search for "older woman"—no need to specify "white"—for references. I question whether you would even wear a mask, and though I have doubts, draw one across your face, like a benevolent voodoo wish for your safety. Yet I can't help but wonder, given our estrangement and my not-quite-white skin, if you saw me today, features obscured by a mask, would you recognize me or be afraid? [End Page 321] Monica Heilman Monica Heilman is an artist and sociology PhD student at Indiana University Bloomington. Her research interests include race and ethnicity, identity, qualitative methodologies, and arts-based research. Monica's current work explores how multiracial identity is shaped by U.S. racial hierarchies. Learn more about her work at monicaheilman.com and artwork at monicaheilmanart.com. Copyright © 2020 Johns Hopkins University Press

Pits or pictures: a comparative study of camera traps and pitfall trapping to survey small mammals and reptiles

Context Camera trapping is a widely used monitoring tool for a broad range of species across most habitat types. Camera trapping has some major advantages over other trapping methods, such as pitfall traps, because cameras can be left in the field for extended periods of time. However, there is still a need to compare traditional trapping methods with newer techniques. Aims To compare trap rates, species richness and community composition of small mammals and reptiles by using passive, unbaited camera traps and pitfall traps. Methods We directly compared pitfall trapping (20-L buried buckets) with downward-facing infrared-camera traps (Reconyx) to survey small reptiles and mammals at 16 sites within a forested habitat in south-western Australia. We compared species captured using each method, as well as the costs associated with each. Key results Overall, we recorded 228 reptiles, 16 mammals and 1 frog across 640 pitfall trap-nights (38.3 animal captures per 100 trap-nights) compared to 271 reptiles and 265 mammals (for species likely to be captured in pitfall traps) across 2572 camera trap nights (20.8 animal captures per 100 trap-nights). When trap effort is taken into account, camera trapping was only 23% as efficient as pitfall trapping for small reptiles (mostly Scincidae), but was five times more efficient for surveying small mammals (Dasyuridae). Comparing only those species that were likely to be captured in pitfall traps, 13 species were recorded by camera trapping compared with 20 species recorded from pitfall trapping; however, we found significant (P&amp;lt;0.001) differences in community composition between the methods. In terms of cost efficacy, camera trapping was the more expensive method for our short, 4-month survey when taking the cost of cameras into consideration. Conclusions Applicability of camera trapping is dependent on the specific aims of the intended research. Camera trapping is beneficial where community responses to ecosystem disturbance are being tested. Live capture of small reptiles via pitfall trapping allows for positive species identification, morphological assessment, and collection of reference photos to help identify species from camera photos. Implications As stand-alone techniques, both survey methods under-represent the available species present in a region. The use of more than one survey method improves the scope of fauna community assessments.

Track plates detect the endangered New Mexico jumping mouse

ABSTRACTThe New Mexico jumping mouse (Zapus luteus luteus, formerly Z. hudsonius luteus), an endangered subspecies found in the southwestern United States, inhabits riparian areas with tall, dense herbaceous vegetation as habitat. To detect presence of this species for use in defining life history and habitat use, we developed and tested 4 noninvasive track‐plate methods, and selected the best for field use. New Mexico jumping mice have unique feet and toes that are readily distinguishable from other small mammals within their geographic range. We created reference photos of rodent tracks that confirmed the unique footprints of the jumping mouse and tested this method against detection with live traps in the Apache–Sitgreaves, Arizona, and Santa Fe, New Mexico, USA, National Forests, 2016 and 2017. When comparing the 2 detection methods, in only 1 of 16 comparisons did results differ, where we captured jumping mice in live traps, but did not detect them with track plates. Based on our success with this approach, we developed a 14‐minute instructional video (https://www.youtube.com/watch?v=i2x0Ydc1XVM) on assembly, deployment, and interpretation of track plates. Although trapping provided specific information about individuals, the noninvasive nature of our track‐plate design minimized risk of injury or mortality to animals and lowered study costs. © 2018 The Wildlife Society.

Strategy for aesthetic photography recommendation via collaborative composition model

In this study, the authors propose a collaborative composition model for automatically recommending suitable positions and poses in the scene of photography taken by amateurs. By analysing aesthetic‐aware features, the authors' strategy jointly takes attention and geometry composition into account to learn the aesthetic manifestation knowledge of professional photographers. Firstly, aesthetic composition representation exploits the strength of visual saliency to explicitly encode the spatial correlation of the professional photos. Secondly, ℓ2 regularised least square is adopted to constrain the representation coefficients, which provides a fast solution in selecting aesthetic candidates collaboratively. In addition, a novel confidence measure scheme is further designed based on reconstruction errors and the reference photos are updated adaptively according to the composition rules. Both qualitative and quantitative evaluations show that the model performs well for the portrait photographing recommendation.

Continuous observation of tree leaf area index at ecosystem scale using upward-pointing digital cameras

Understanding the dynamics of leaf area index (LAI) at ecosystem scale is critical for evaluating and modeling the response of vegetation to environmental variability and change. We present an approach for quantifying daily ecosystem LAI that integrates automatic acquisition of a small number of reference photos, viewed towards the zenith, and a larger number of photos covering a spatially extensive area that were manually acquired during periodic field trips. We present an image archive spanning three years for an oak-savanna ecosystem in California to identify the timing of phenological and disturbance events, and to quantify the seasonal to interannual variability of tree LAI at ecosystem scale. The digital camera-derived LAI, corrected for clumping effects, agreed well (r2=0.94, root mean square error=0.05) with independent estimates of LAI from litterfall traps and the LAI-2000 Plant Canopy Analyzer. Using the Moderate Resolution Imaging Spectroradiometer (MODIS) LAI product, which combined grass and tree LAI, we obtained realistic seasonal patterns of ecosystem LAI that corresponded with tree LAI from the digital cameras for the period when grass was dead. The digital camera method is an easy and inexpensive way to monitor LAI at ecosystem scale. This method can be used for testing and improving phenology models, evaluating remote-sensing-derived LAI products, and quantifying forest structures in rapidly changing environments.

THE USE OF COMPUTED TOMOGRAPHIC THREE-DIMENSIONAL RECONSTRUCTIONS TO DEVELOP INSTRUCTIONAL MODELS FOR EQUINE PELVIC ULTRASONOGRAPHY

Ultrasonography has gained increased utility to diagnose pelvic fractures in horses; however, internal pelvic contours can be difficult to appreciate from external palpable landmarks. We developed three-dimensional (3D) simulations of the pelvic ultrasonographic examination to assist with translation of pelvic contours into two-dimensional (2D) images. Contiguous 1mm transverse computed tomography (CT) images were acquired through an equine femur and hemipelvis using a single slice helical scanner. 3D surface models were created using a DICOM reader and imported into a 3D modeling and animation program. The bone models were combined with a purchased 3D horse model and the skin made translucent to visualize pelvic surface contours. 3D models of ultrasound transducers were made from reference photos, and a thin sector shape was created to depict the ultrasound beam. Ultrasonographic examinations were simulated by moving transducers on the skin surface and rectally to produce images of pelvic structures. Camera angles were manipulated to best illustrate the transducer-beam-bone interface. Fractures were created in multiple configurations. Animations were exported as QuickTime movie files for use in presentations coupled with corresponding ultrasound videoclips. 3D models provide a link between ultrasonographic technique and image generation by depicting the interaction of the transducer, ultrasound beam, and structure of interest. The horse model was important to facilitate understanding of the location of pelvic structures relative to the skin surface. While CT acquisition time was brief, manipulation within the 3D software program was time intensive. Results were worthwhile from an instructional standpoint based on user feedback.

Validation of the use of digital photography for estimating portion sizes

There is no consensus and the data are scarce on the best method for assessing diets of young children. The purpose of this study was to validate the use of digital photography as a dietary assessment method for use with preschool children. Researchers prepared and photographed 23 foods presented as ten versions of six meals. Investigators selected foods used in the estimations from a Head Start preschool lunch menu. As researchers added individual foods to a plate, they weighed each food portion to determine actual weight. From photographs on the computer, five raters estimated portion sizes for the servings of food, using a library of reference photos representing each food in 10‐gram increments. Raters viewed the meal photos and selected what they believed to be the corresponding portion size reference photo for each food on the plate. The average correlation between estimated weights across raters was 0.65. The mean difference in actual versus estimated weight of food was 1.92 grams across raters. This small mean gram difference suggests that the difference between actual and estimated nutrient analysis of foods would be very small. Thus, this method is a valid approach to collecting nutrient intake data. In the future, researchers will validate the use of this method for application in community settings including Head Start classrooms and participants’ homes.Funded by NIH grants T32 HD007445 and R01 CA107545

Anterior Segment Measurements Using Digital Photography: A Simple Technique

The purpose of this study is to describe a simple method of measuring anterior segment lesions from digital slit lamp images and confirm reliability of the technique. Ten reference photos were taken of a PD ruler, refocusing on the ruler for each photo. Using Adobe Photoshop, the number of pixels per millimeter squared (area) and 1 mm (linear) were recorded for each photo. An Excel spreadsheet was set up to convert pixels to millimeters and millimeters squared. Interrater reliability was determined by two observers who independently calculated the area of 69 epithelial defects. A Bland and Altman plot was used to demonstrate the agreement between the two doctors. Interrater reliability was excellent as measured by an intraclass correlation coefficient (ICC 2,1) = 0.99. From the Bland and Altman plot it was determined that in 95% of cases, the area of the epithelial defect measured by provider 2 may be as much as 1.17 mm less than or 1.09 mm greater than that measured by provider 1. This simple method, which allows accurate measurements from digital images using common off-the-shelf software, is a valuable tool for clinical documentation as well as for research purposes.

Within country accuracy of tree crown transparency estimates using the image analysis system CROCO: a case study from Switzerland

We evaluated the accuracy of tree crown transparency estimates made with the image analysis system CROCO compared to visual estimates by field observers in Switzerland. Data from Swiss field training courses in 1998 and 1999 each comprising four sessions were used and the median values of the different observers’ scores were assumed as the true values. First, Swiss reference photographs alone were used to obtain the calibration curves needed to predict crown transparency from the image analysis measure DSO. In this case, the CROCO estimates were biased for many species. Second, data from the first three field training sessions were added to the reference photos to derive the calibration curves. In this second case, CROCO provided unbiased estimates of both years, while the estimates by some of the field teams were significantly different from the true values. The result suggests that CROCO can be used as a more reliable reference to detect observer bias within a country than control observers. We conclude that few reference photographs alone are not sufficient, so additional field data should also be used to estimate crown transparency more accurately using CROCO.