Photographic Detection Research Articles

Accuracy of Artificial Intelligence-Based Photographic Detection of Gingivitis

ObjectivesGingivitis is one of the most prevalent plaque-initiated dental diseases globally. It is challenging to maintain satisfactory plaque control without continuous professional advice. Artificial intelligence may be used to provide automated visual plaque control advice based on intraoral photographs. MethodsFrontal view intraoral photographs fulfilling selection criteria were collected. Along the gingival margin, the gingival conditions of individual sites were labelled as healthy, diseased, or questionable. Photographs were randomly assigned as training or validation datasets. Training datasets were input into a novel artificial intelligence system and its accuracy in detection of gingivitis including sensitivity, specificity, and mean intersection-over-union were analysed using validation dataset. The accuracy was reported according to STARD-2015 statement. ResultsA total of 567 intraoral photographs were collected and labelled, of which 80% were used for training and 20% for validation. Regarding training datasets, there were total 113,745,208 pixels with 9,270,413; 5,711,027; and 4,596,612 pixels were labelled as healthy, diseased, and questionable respectively. Regarding validation datasets, there were 28,319,607 pixels with 1,732,031; 1,866,104; and 1,116,493 pixels were labelled as healthy, diseased, and questionable, respectively. AI correctly predicted 1,114,623 healthy and 1,183,718 diseased pixels with sensitivity of 0.92 and specificity of 0.94. The mean intersection-over-union of the system was 0.60 and above the commonly accepted threshold of 0.50. ConclusionsArtificial intelligence could identify specific sites with and without gingival inflammation, with high sensitivity and high specificity that are on par with visual examination by human dentist. This system may be used for monitoring of the effectiveness of patients’ plaque control.

Status and distribution of jaguarundi in Texas and Northeastern México: Making the case for extirpation and initiation of recovery in the United States.

The jaguarundi (Puma yagouaroundi) is a small felid with a historical range from central Argentina through southern Texas. Information on the current distribution of this reclusive species is needed to inform recovery strategies in the United States where its last record was in 1986 in Texas. From 2003 to 2021, we conducted camera‐trap surveys across southern Texas and northern Tamaulipas, México to survey for medium‐sized wild cats (i.e., ocelots [Leopardus pardalis], bobcats [Lynx rufus], and jaguarundi). After 350,366 trap nights at 685 camera sites, we did not detect jaguarundis at 16 properties or along 2 highways (1050 km2) in Texas. However, we recorded 126 jaguarundi photographic detections in 15,784 trap nights on 2 properties (125.3 km2) in the northern Sierra of Tamaulipas, Tamaulipas, México. On these properties, latency to detection was 72 trap nights, with a 0.05 probability of detection per day and 0.73 photographic event rate every 100 trap nights. Due to a lack of confirmed class I sightings (e.g., specimen, photograph) in the 18 years of this study, and no other class I observations since 1986 in the United States, we conclude that the jaguarundi is likely extirpated from the United States. Based on survey effort and results from México, we would have expected to detect jaguarundis over the course of the study if still extant in Texas. We recommend that state and federal agencies consider jaguarundis as extirpated from the United States and initiate recovery actions as mandated in the federal jaguarundi recovery plan. These recovery actions include identification of suitable habitat in Texas, identification of robust populations in México, and re‐introduction of the jaguarundi to Texas.

A novel colorimetric sensor based onmodified mesoporous silica nanoparticles for rapid on-site detection of nitrite.

A novel colorimetric sheet based on Griess reagent-doped mesoporous silica nanoparticles was developed for nitrite detection. Griess reagent was adsorbed on long-range ordered hexagonal mesoporous silica nanoparticles and developed ink-bottle pores with some disorder. When the modified nanoparticles were bound using starch to fabricate a thin (~ 313μm) colorimetric sheet, spherical particles with a rougher surface and some distortion of their mesoporosity were observed. The sheet was used in conjunction with digital image colorimetry (DIC) andprovides a wide linear range of0.05 to 2.50mgL-1 with a low detection limit (15.0μgL-1-NO2-, equal to 4.5μgL-1 NO2--N), good inter-day precision (1.93%RSD), and excellent precision (2.67% relative error). The colorimetric sensors produced from the sheet costs only 0.04 USD each, while the DIC uses a standard smartphone for photographic detection. The method developed offers an easier and cheaper means of conducting rapid on-site determinationof nitrite in water with reliable quantitative results. Graphical abstract.

Colorimetric Method for the Determination of Proteins Using Immobilized Microbial Protease and a Smartphone Camera

The determination of proteins is very important in industry, medicine, analytical chemistry, and biotechnology. A novel colorimetric method using magnetic microparticles as a carrier of immobilized microbial protease on surface was developed and smartphone camera was employed for detection. Folin and Ciocalteu’s phenol reagent was selected as the chromogen and casein from bovine milk was used as the substrate. Casein was cleaved by immobilized enzyme to tyrosine, which together with chromogen spontaneously generated a blue color measurable by a spectrophotometer or photographic detection by a smartphone camera. The photographs were processed by a computer and selected points were transformed into RGB numeric values. A reference spectrophotometric method provided results that provide a Michaelis–Menten dependence with the Michaelis–Menten constant equal to 0.11 mmol/l and a limit of detection equal to 41 µg/ml. The novel method using spectrophotometric detection shows the same Michaelis–Menten dependence with a Michaelis–Menten constant equal to 0.11 mmol/l with a limit of detection of 4.27 µg/ml. The RGB method using a smartphone for the determination of proteins was demonstrated to be an alternative to the standard spectrophotometer. The limit of detection was 242 µg/ml, which is sufficient for protein determination in real samples. Easy processing, good analytical results, and low demands on equipment make this method a suitable tool for a wide range of applications in industry, analytical chemistry, medicine, biotechnology, and biosensors.

Colorimetric and photographic detection of bacteria in drinking water by using 4-mercaptophenylboronic acid functionalized AuNPs

Water-borne bacterial infection is a growing problem worldwide. Rapid and reliable contamination detection method for drinking-water is in high demand. Herein, a colorimetric bacteria detection strategy based on 4-mercaptophenylboronic acid (4-MPBA) functionalized gold nanoparticles (AuNPs) is introduced. The immobilization of 4-MPBA-AuNPs on the surface of bacteria cells via covalent bonds impedes NaCl induced aggregation of 4-MPBA-AuNPs. The color shift due to AuNPs aggregation is observable with the naked eye and quantifiable through UV–Vis absorbance spectrum or digital-camera-based RGB analysis. This platform is validated with five strains of bacteria, including gram-positive and gram-negative strains, which demonstrated its capability for broad-spectrum bacteria detection. By using UV–Vis absorbance spectrum for bacteria quantification, a linear relationship is observed between absorbance intensity at 524 nm and E. coli concentration from 104-107 CFU/mL (R2 = 0.9577). This method only requires the addition of 10 μL detection sample and 10 μL NaCl, and the results are obtainable within 20 min. This study is the first documented report to utilize bacterial inhibition of salt-induced aggregation of 4-MPBA-AuNPs for broad-spectrum bacteria detection. The proposed colorimetric method offers a cost-effective solution for drinking-water quality control.

Acoustic and camera surveys inform models of current and future vertebrate distributions in a changing desert ecosystem

AbstractAimMaintaining biodiversity in the face of land use and climate change is a paramount challenge, particularly when distributions of many species remain incompletely known. Emerging technologies help address this data deficiency by facilitating the collection of spatially explicit data for multiple species from multiple taxa. In this study, we combine acoustic and visual sensor surveys to inform conservation and land use planning in an area experiencing rapid climate and land use change.LocationMojave Desert, California, United States.MethodsWe deployed camera traps and acoustic detectors at 210 sites between March and July 2016. We identified photographic detections of mammals and acoustic recordings of songbirds to the species level and used multispecies occupancy models to estimate and evaluate species' occupancy probabilities. We then extrapolated model results to the region and forecasted how projected climate and land use changes might affect species' occupancy probabilities in 50 years. Lastly, we identified areas with high conservation value (i.e., high relative species richness) now and in 50 years, and related the distributions of these areas to land use designations.ResultsWe detected 15 mammal and 68 songbird species. At the community level, occupancy decreased with increasing temperatures and distances to woodlands. We forecasted that occupancy probabilities and areas with high conservation value would decline in 50 years due to projected increases in maximum temperatures and identified that up to 43%, 24% and 27% of land designated for renewable energy development, recreation and military activities, respectively, encompassed these high value areas.Main conclusionsCooler areas close to woodlands and water are of high conservation value to mammals and songbirds in the Mojave. These areas will become increasingly limited with changing climate, however, making their protection from human disturbance imperative. We encourage continued use of visual and acoustic sensors across large spatial, temporal and taxonomic scales as tools to inform land use and wildlife conservation.

Less invasive monitoring of cougars in Colorado's front range

ABSTRACTFrom 2014 to 2016, in the Front Range of Colorado, USA, we assessed noninvasive approaches to sampling cougar (Puma concolor) populations in an attempt to provide a new method that would be less field intensive, less expensive, and could be applied over large spatial extents compared with current methods. We assessed the use of predator calls to lure cougars to a site with remote camera traps for detection and also evaluated hair snags at sites to noninvasively identify individual animals. Predator calls effectively attracted cougars to specific sites with an average of 82 unique photographic detections of cougars per survey year (0.03 detections/trap‐night). However, obtaining hair samples from these animals was less effective because animals did not always pass through hair snags and ability to uniquely identify individuals by genotype was poor. We evaluated different approaches to estimating cougar density and found mark–resight to be a viable option in our study system. Mark–resight density estimate after correcting for partial use of the sampling area by cougars was 4.1 cougars/100 km2 (95% CI = 2.4, 5.8). Our results indicate that combining methods of noninvasive genetic sampling and auditory calls to monitor cougar populations can provide reliable density estimates over large geographic areas and areas with significant amounts of inaccessible private lands. © 2019 The Wildlife Society.

Small cat surveys: 10 years of data from Central Kalimantan, Indonesian Borneo

We present an update on the photographic detections from camera traps and the activity patterns of Borneo’s four small cats, namely, Sunda Leopard Cat Prionailurus javanensis, Flat-headed Cat P. planiceps, Marbled Cat Pardofelis marmorata, and Bay Cat Catopuma badia, at two sites in Central Kalimantan, Indonesia. Camera trap survey data of 10 years (2008–2018) from the first site in Sebangau provide details about the temporal partitioning of these small cats from each other but overlap with Sunda Clouded Leopard Neofelis diardi. The activity of Flat-headed Cat was higher after midnight and that of Leopard Cat at night with no clear preference before or after midnight. The Marbled Cat is predominantly diurnal, but the remaining three cats have flexible activity periods. While limited data are available from Rungan, the second site, we confirmed the presence of all four small cat species found on Borneo, though we have insufficient data to comment on the Bay Cat. The cat sightings, however, are intermittent and may reflect the unprotected status of this forest. Leopard Cats appear relatively unaffected by habitat disturbance based on encounter rates on camera traps. Conservationists, both NGOs and the government, must pay particular attention to specialists like Flat-headed Cats and Bay Cats when assessing habitat suitability for long-term cat conservation.

Citizen science data facilitate monitoring of rare large carnivores in remote montane landscapes

Population monitoring of large carnivores, particularly in remote montane landscapes, represents a considerable conservation challenge. Occupancy modeling using repeated detection/non-detection surveys offers a practical and robust tool for assessments of this type. Sign surveys or photographic detections have been the two primary survey methods to inform occupancy models. However, these approaches are expensive to implement and resource-intensive. Thus, their applicability for assessing the distribution of rare large carnivores residing in inaccessible landscapes is limited, particularly when large scale species monitoring is desired. Here, our intent was to predict the occupancy of the endangered Persian leopard (Panthera pardus saxicolor) inhabiting the rugged mountains of northeastern Iran. Using a Bayesian occupancy modeling framework, we compared patterns of leopard occupancy derived from standardized monitoring data implemented using spatially-replicated sign surveys to those informed by citizen scientist observations. We found that leopard occupancy probability was comparable between the two survey methods (sign survey = 0.92, 95% CI 0.85–1 and citizen science = 0.94, 95% CI 0.88–1) though detection probability varied (sign survey = 0.52, 95% CI 0.46–0.58 and citizen science = 0.25, 95% CI 0.18–0.32). The magnitude of effect among the environmental covariates that predicted leopard occupancy probability was also similar for the two methods. Thus, while yielding comparable predictions, the citizen science approaches were half the cost of sign surveys. The implementation of the effective citizen scientist data enabled us to expand by two-fold the monitored area while halving the costs in comparison to the area investigated via sign surveys. Our paper demonstrates that citizen science surveys represent a cost-effective, reliable, and surprisingly overlooked means to efficiently assess occupancy, particularly for rare large carnivores inhabiting mountainous landscapes.

Predicting the Potential Distribution of the Sierra Nevada Red Fox in the Oregon Cascades

Abstract The Sierra Nevada red fox Vulpes vulpes necator is a native subspecies associated with subalpine regions in the Sierra Nevada and Cascade mountain ranges of California and Oregon. In the past century, the Sierra Nevada red fox experienced a major range contraction and decline in California. However, the number, size, and connectivity of populations extant in Oregon remain unclear. This knowledge gap impedes efficient monitoring and hinders development of a cohesive conservation strategy for the subspecies. The historical range is large and includes rugged terrain with low accessibility; therefore, a predictive model is needed to facilitate more comprehensive and systematic surveys in the future. We initiated a multiagency collaborative effort to survey portions of the range in the Oregon Cascades during 2011–2016 (verified genetic and photographic detections) and to assemble existing sighting reports dating back to 1985 (unverified), which we used to create Maxent models to predict the potential distribution of Sierra Nevada red fox within Oregon. To identify optimal levels of model complexity, we compared cross-validation accuracy of models that varied in levels of protection against overfitting (regularization). The highest-performing models utilized intermediate regularization, and included minimum January temperature and land-cover type. Regardless of regularization or data set (verified detections, all putative detections), all models agreed in predictions of a high-probability region covering approximately 3,470 km2 or 6% of the Cascade region, corresponding to the high-elevation portion of the crest. With the exception of a gap between Mount Hood and Mt. Jefferson, this core area of predicted presence was continuous along the north–south extent of the crest, suggesting a capacity for high connectivity among observed clusters of occurrence. Use of modeled potential distributions in future survey design will improve efficiency of field data collection, facilitating more precise evaluations of the distribution, abundance, and genetic integrity and connectivity of Sierra Nevada red fox in Oregon.

Carnivore distributions in Botswana are shaped by resource availability and intraguild species

AbstractThe composition of ecological communities is shaped by the interplay between interspecific interactions and species’ habitat and food requirements. The influence of interspecific interactions is particularly widespread within carnivore guilds, where species must balance the trade‐off between resource acquisition and potentially costly encounters with intraguild competitors/predators. We tested whether intraguild species activity or resource availability had a stronger influence on the seasonal distributions of 10 carnivore species in northern Botswana. We deployed 132 camera stations across a 330 km2 area during the 2014 dry season and 2015 wet season. For each species and season, we developed occupancy models based on resource availability (i.e. prey, vegetation and water) and on intraguild species (i.e. photographic detection rates of intraguild competitors and predators). We then used k‐fold cross‐validation to assess the relative predictive ability of each model. Carnivore distributions were generally negatively associated with dense vegetation and contrary to expectations, positively associated with the detection rate of intraguild species. This suggests competitor/predator avoidance did not play a large role in influencing carnivore distributions in northern Botswana, a result that differs from other systems, and might be attributed to differences in habitat, carnivore densities, and prey availability. The predictive ability of our resource availability vs. intraguild species models differed between seasons and among species. Leopard distributions, for example, were best predicted by resource availability during the dry season and by the activity of intraguild species in the wet season. The majority of seasonal distributions were best predicted by intraguild species activity or a combination of both resource availability and intraguild species activity. As environments continually change, studies similar to ours are pertinent as they can be used to monitor distributions of wildlife communities and to better understand the relative importance of the diversity of ecological processes impacting wildlife communities.

Enzyme-Based Test Strips for Visual or Photographic Detection and Quantitation of Gaseous Sulfur Mustard.

Sulfur mustard is a chemical agent of high military and terroristic significance. No effective antidote exists, and sulfur mustard can be fairly easily produced in large quantity. Rapid field testing of sulfur mustard is highly desirable. Existing analytical devices for its detection are available but can suffer from low selectivity, laborious sample preparation, and/or the need for complex instrumentation. We describe a new kind of test strip for rapid detection of gaseous sulfur mustard that is based on its degradation by the enzyme haloalkane dehalogenase that is accompanied by a change of local pH. This change can be detected using pH indicators contained in the strips whose color changes from blue-green to yellow within 10 min. In addition to visual read-out, we also demonstrate quantitative reflectometric readout by using a conventional digital camera based on red-green-blue data acquisition. Organic haloalkanes, such as 1,2-dichloroethane, have a negligible interfering effect. The visual limit of detection is 20 μg/L, and the one for red-green-blue read-out is as low as 3 μg/L. The assays have good reproducibility ±6% and ±2% for interday assays and intraday assays, respectively. The strips can be stored for at least 6 months without loss of function. They are disposable and can be produced fairly rapidly and at low costs. Hence, they represent a promising tool for in-field detection of sulfur mustard.

The First Estimates of Marbled Cat Pardofelis marmorata Population Density from Bornean Primary and Selectively Logged Forest.

The marbled cat Pardofelis marmorata is a poorly known wild cat that has a broad distribution across much of the Indomalayan ecorealm. This felid is thought to exist at low population densities throughout its range, yet no estimates of its abundance exist, hampering assessment of its conservation status. To investigate the distribution and abundance of marbled cats we conducted intensive, felid-focused camera trap surveys of eight forest areas and two oil palm plantations in Sabah, Malaysian Borneo. Study sites were broadly representative of the range of habitat types and the gradient of anthropogenic disturbance and fragmentation present in contemporary Sabah. We recorded marbled cats from all forest study areas apart from a small, relatively isolated forest patch, although photographic detection frequency varied greatly between areas. No marbled cats were recorded within the plantations, but a single individual was recorded walking along the forest/plantation boundary. We collected sufficient numbers of marbled cat photographic captures at three study areas to permit density estimation based on spatially explicit capture-recapture analyses. Estimates of population density from the primary, lowland Danum Valley Conservation Area and primary upland, Tawau Hills Park, were 19.57 (SD: 8.36) and 7.10 (SD: 1.90) individuals per 100 km2, respectively, and the selectively logged, lowland Tabin Wildlife Reserve yielded an estimated density of 10.45 (SD: 3.38) individuals per 100 km2. The low detection frequencies recorded in our other survey sites and from published studies elsewhere in its range, and the absence of previous density estimates for this felid suggest that our density estimates may be from the higher end of their abundance spectrum. We provide recommendations for future marbled cat survey approaches.

Fast collective photographic subject detection without pixels by an assumption about a shoot and its elevation angle

In this paper a method that discovers socially attracted photographic subjects in real time is proposed. This needs only non-pixel information of a digital photograph for this discovery which include a time of shoot, GPS location where a photograph is taken, a bearing of the shoot and so on. Thinformation is sufficiently lightweight so that we expect that a congestion-free photo sharing wireless network service can be achieved. In this paper, a photography behavior model which considers an elevation angle of a shoot makes the accuracy better.

Photographic Detection of Scattering Light of Suspended Solids in the Sea

Photographic Detection of Scattering Light of Suspended Solids in the Sea

Photographic Detection of Cadmium(II) and Zinc(II) Ions

The binding properties of two 2,8-dithia-5-aza-2,6-pyridinophane-based ligands, bearing respectively coumarin (HNCum) and naphthol-benzoxaazole (HNBO) optically active units as pendant arms, were studied with the aim of transition metal cations selective detection. The novel all-solid-state optodes were obtained by inclusion of fluorophores inside PVC-based polymeric films. The influence of lipophilic sites incorporation and plasticizer nature on pyridinophane-based optodes response was investigated. An enhanced selectivity of the ligands towards Cadmium(II) and Zinc(II) ions was detected. Light Emitting Diode, LED 380nm, was used as light source, and a digital camera as a signal detector. The results obtained demonstrated the suitability of the developed optodes to perform fast and low cost simultaneous monitoring of Zn2+and Cd2+ ions by means of familiar devices and chemometrics.

Genetic and Photographic Detections Document Sierra Nevada Red Fox in the Northern Cascades of Oregon

Abstract The red fox (Vulpes vulpes) exhibits a high degree of ecological plasticity, but the indigenous Sierra Nevada red fox (SNRF; V. v. necator) belongs to an evolutionarily divergent complex of montane red foxes, apparently specialized to subalpine habitats. Mountain red foxes historically occurred among isolated sky islands of the major mountain ranges of the western U.S., with the SNRF occurring along the Pacific Crest of Oregon and California. Little is known about the current distribution and status of SNRF in Oregon, which complicates regulatory decisions and impedes conservation of the subspecies. We conducted a survey in the northern Oregon Cascade Range during 2012–2014 using baited ground-level camera stations and noninvasive genetic sampling. Our objectives were to document red fox occurrences in our study area and investigate their genetic affinities using mitochondrial DNA sequences to indicate maternal haplotype. We detected red foxes at 11 of 41 camera stations and collected DNA from 18...

Probable photographic detection of the natural seventh-order rainbow.

We present a stacked and contrast-enhanced image comprised of 12 digital photographs that shows a series of color hues in the correct order and location to be part of the seventh-order rainbow. The observation was made on September 22, 2013, near Magdalena in New Mexico (USA). The seventh-order rainbow is located at 64° from the Sun in a region of the sky with little interference from the zero-order glow. The color hues in the image range from red to blue-violet, spanning about 12° in total extent; their locations generally agree with a numerical Debye-series simulation of the seventh-order rainbow. Despite the low color contrast of the seventh-order rainbow, the current observation indicates that it is feasible with current digital-imaging technology to detect this higher-order rainbow in near-ideal atmospheric conditions.

New variable stars on digitized plates of the Moscow collection. Field SA9

The new stage of work on digitizing the astronomical plates of the Sternberg Astronomical Institute’s collection and searching for new variable stars using the digitized photographic plates includes a considerable improvement of the automated search techniques for objects with variable brightnesses. In particular, the technique for taking into account the non-linear response of the photographic light detector has been improved. Applying the improved techniques to 182 digitized images of a 10° × 10° field centered at SA9, obtained from scanning photographic plates taken with the Sternberg Astronomical Institute 40-cm astrograph, has enabled the discover and study of 77 new variable stars (MDV 519–595). These include 3 Cepheids, 2 probable BY Draconis stars, 65 eclipsing binaries, 3RR Lyrae stars, 1 high-amplitude δ Scuti star (HADS), and 3 irregular variables. Special CCD observations have confirmed the presence of brightness variations in 7 of the 77 variables that were initially considered uncertain.

Wildcat population density on theEtna volcano,Italy: a comparison of density estimation methods

AbstractTheEuropean wildcat is an elusive felid that is declining across its range.Sicily hosts a distinctive insular wildcat population, the conservation of which requires much better ecological knowledge than is currently available, particularly population density. We simultaneously used two noninvasive methods (camera‐trapping and scat‐collection) to estimate the population density of wildcats on theEtna volcano. We conducted genetic analyses to identify individuals and to detect potential hybridization with the domestic cat. We analyzed individual capture‐histories from camera‐trapping and scat‐collection using the spatially explicit capture‐recapture (SECR) model. Furthermore, we applied the random encounter model (REM), which does not require individual identification, to the camera‐trapping data. We identified 14 wildcats from 70 photographic detections (6.48 detections/100 trap‐days) obtained from 1080 camera‐trapping days over 4 months, and we estimated to have identified all the individuals living in the study area (10.9 km−2). On the contrary, we identified 10 wildcats from 14 out of 39 scats collected from 391 km of transects walked. The estimated densities (individuals km−2 ± se) were 0.32 ± 0.1 (SECRcamera‐trapping), 1.36 ± 0.73 (SECRscat‐collection) and 0.39 ± 0.03 (REM). The population density estimates obtained fromSECRcamera‐trapping andREMoverlapped, although we recommend care when applying the latter. TheSECRscat‐collection gave the highest population density (and less precise) estimates because of the low number of capture and recaptures; however, the population size estimated with this method matched the number of individuals photographed. The population density of the wildcat inEtna falls in the medium‐high range of those reported in literature, highlighting the role of this ecosystem for the long‐term conservation of the wildcat inSicily. Camera‐trapping is confirmed as a useful tool to assess the wildcat population density and, in this case, was complemented by the genetic analysis that confirmed individual identity.