Find research datasets worth reusing

opencc-zeroNov 2022View details →

Northern Nevada wildlife and topography: Camera trapping data set for 14 mammal species collected from 100 sampling sites in northwestern Nevada

Camera traps are one of the most common field techniques for surverying terrestrial mammal communities and thus, much work has gone into understanding how different factors influence species detection at camera trap locations. However, the effect of fine-scale topography, such as terrain slope and position, on wildlife detection has not been explicitly quantified despite strong effects of topography on animal movement in mountainous regions. This data set contains weekly detection non-detection data for 14 mammal species from 100 camera traps sites monitored for 28 months (June 2018 - September 2020) in northwestern Nevada, U.S.A. This sampling extent was split into 3 month sampling seasons, exclusive of winter (Dec, Jan, Feb) and spring 2020, when data were sparse. In addition to species detection data, that dataset includes topographic variables at cameras sites: 1) terrain slope, calculated in R package raster from a 10m digital elevation model and 2) Topographic position index averaged across three buffer sizes around points 270m, 810m, and 2430m. The land cover variables proportion mixed conifer and proportion pinyon-juniper woodland within a 5000m buffer of sites are also included. Both are derived from the USDA/US DOI Landfire 2016 dataset. The luring variable indicates whether attractant was applied at a site during a given week, the effect of which was assumed to last for a month after the last application.

opencc-zeroDec 2022View details →

Inferring predator-prey interactions from camera traps: A Bayesian co-abundance modelling approach

Predator-prey dynamics are a fundamental part of ecology, but directly studying interactions has proven difficult. The proliferation of camera trapping has enabled the collection of large datasets on wildlife, but researchers face hurdles inferring interactions from observational data. Recent advances in hierarchical co-abundance models infer species interactions while accounting for two species' detection probabilities, shared responses to environmental covariates, and propagate uncertainty throughout the entire modelling process. However, current approaches remain unsuitable for interacting species whose natural densities differ by an order of magnitude and have contrasting detection probabilities, such as predator-prey interactions, which introduce zero-inflation and overdispersion in count histories. Here we developed a Bayesian hierarchical N-mixture co-abundance model that is suitable for inferring predator-prey interactions. We accounted for excessive zeros in count histories using an informed zero-inflated Poisson distribution in the abundance formula and accounted for overdispersion in count histories by including a random effect per sampling unit and sampling occasion in the detection probability formula. We demonstrate that models with these modifications outperform alternative approaches, improve model goodness-of-fit, and overcome parameter convergence failures. We highlight its utility using 20 camera trapping datasets from 10 tropical forest landscapes in Southeast Asia and estimate four predator-prey relationships between tigers, clouded leopards, and muntjac and sambar deer. Tigers had a negative effect on muntjac abundance, providing support for top-down regulation, while clouded leopards had a positive effect on muntjac and sambar deer, likely driven by shared responses to unmodelled covariates like hunting. This Bayesian co-abundance modelling approach to quantify predator-prey relationships is widely applicable across species, ecosystems, and sampling approaches, and may be useful in forecasting cascading impacts following widespread predator declines. Taken together, this approach facilitates a nuanced and mechanistic understanding of food-web ecology.

opencc-zeroDec 2022View details →

Large-antlered muntjac (Muntiacus vuquangensis) camera-trap photos from Virachey NP, Cambodia

We present evidence of scent marking in the large-antlered muntjac (Muntiacus vuquangensis). Given the importance of scent marking in individual recognition among ungulates, this behavior may serve to communicate the fitness cost of antagonistic interactions among rival males and could serve as a mechanism for mate assessment among females.

opencc-zeroJan 2023View details →

A novel camera trapping method for individually identifying pumas by facial features

Camera traps (CTs), used in conjunction with capture-mark-recapture analyses (CMR; photo-CMR), are a valuable tool for estimating abundances of rare and elusive wildlife. However, a critical requirement of photo-CMR is that individuals are identifiable in CT images (photo-ID). Thus, photo-CMR is generally limited to species with conspicuous pelage patterns (e.g., stripes or spots) using lateral-view images from CTs stationed along travel paths. Pumas (Puma concolor) are an elusive species for which CTs are highly effective at collecting image data, but their suitability to photo-ID is controversial due to their lack of pelage markings. For a wide range of taxa, facial features are useful for photo-ID, but this method has generally been limited to images collected with traditional handheld cameras. Here we evaluate the feasibility of using puma facial features for photo-ID in a CT framework. We consider two issues: 1) the ability to capture puma facial images using CTs, and 2) whether facial images improve human ability to photo-ID pumas. We tested a novel CT accessory that used light and sound to attract the attention of pumas, thereby collecting face images for use in photo-ID. Face captures rates increased at CTs that included the accessory (n = 208, χ2 = 43.23, P ≤ 0.001). To evaluate if puma faces improve photo-ID, we measured the inter-rater agreement of 5 independent assessments of photo-ID for 16 of our puma face capture events. Agreement was moderate to good (Fleiss' kappa = 0.54, 95% CI = 0.48–0.60), and was 92.90% greater than a previously published kappa using conventional CT methods. This study is the first time such a technique has been used for photo-ID, and we believe a promising demonstration of how photo-ID may be feasible for an elusive but unmarked species.

Fracture propagation captured by high-speed camera 2: Video (configuration 4)

The video presents the target plate captured by high-speed camera 2 during testing, and shows the damage for configuration 4. A line can be observed between the first and the third impact and a lighter one between the second and third impact. A deeper analysis of the three enlarged holes highlights the location of crack tips and the crack between the first and the third hole. The line between impact 1 and 3 corresponds, to a full crack formation on the plate. On the contrary, there is only an initiation of crack formation between impacts 2 and 3.

Penetration and perforation process of 12.7mm FSP on plate captured by high-speed camera 2: Video (Test 2 of configuration 2).

The penetration and the perforation process of configuration 2 is recorded using HSC 2. It is shown that at the moment of impact, the FSP causes a localized bulge at the rear side of the plate. A localized shearing of material, induced by the FSP front surface, is created in the contact zone. Immediately after, the FSP penetrates and a circular plug is punched through the plate, leaving a clean cut hole. Subsequently, the punched plug can be seen attached to the front surface of the projectile. After that, the punched plug is separated from the FSP.

"Pitch-angle of FSP before and after ballistic impact on plate captured by high-speed camera 1: Video (configuration 2)".

The figure reveals that the pitch-angle of a 12.7mm FSP both before and after the ballistic impact on an Aluminum plate is low.

opencc-by-4.0Mar 2023View details →

Image dataset for training of an insect detection model for the Insect Detect DIY camera trap

This dataset contains images of an artifical flower platform with different insects sitting on it or flying above it. All images were automatically recorded with the <a href="https://maxsitt.github.io/insect-detect-docs/">Insect Detect DIY camera trap</a>, a hardware combination of the Luxonis OAK-1, Raspberry Pi Zero 2 W and PiJuice Zero pHAT for automated insect monitoring (<a href="https://doi.org/10.1101/2023.12.05.570242">bioRxiv preprint</a>).<h2>Classes</h2>The following object classes were annotated in this dataset:<ul><li>wasp (mostly Vespula sp.)</li><li>hbee (Apis mellifera)</li><li>fly (mostly Brachycera)</li><li>hovfly (various Syrphidae, e.g. Episyrphus balteatus)</li><li>other (all Arthropods with insufficient occurences, e.g. various Hymenoptera, true bugs, beetles)</li><li>shadow (shadows of the recorded insects)</li></ul>View the <a href="https://universe.roboflow.com/maximilian-sittinger/insect_detect_detection/health">Health Check</a> for more info on class balance.<h2>Versions</h2><ul><li><a href="https://universe.roboflow.com/maximilian-sittinger/insect_detect_detection/dataset/4">v4 insect_detect_416_1class</a><ul><li>squashed to square (aspect ratio 1:1)</li><li>downscaled to 416x416 pixel</li><li>all classes merged into one class ("insect")</li></ul></li><li><a href="https://universe.roboflow.com/maximilian-sittinger/insect_detect_detection/dataset/5">v5 insect_detect_raw_4K</a><ul><li>original images in 4K resolution (3840x2160 pixel)</li></ul></li><li><a href="https://universe.roboflow.com/maximilian-sittinger/insect_detect_detection/dataset/7">v7 insect_detect_320_1class</a><ul><li>squashed to square (aspect ratio 1:1)</li><li>downscaled to 320x320 pixel</li><li>all classes merged into one class ("insect")</li></ul></li></ul><h2>Deployment</h2>You can use this dataset as starting point to train your own insect detection models. Check the <a href="https://maxsitt.github.io/insect-detect-docs/modeltraining/train_detection/">model training instructions</a> for more information.Open source Python scripts to deploy the trained models can be found at the <a href="https://github.com/maxsitt/insect-detect">insect-detect GitHub repo</a>.

opencc-by-4.0Mar 2023View details →

Intersection Monitoring: A dataset for vehicle detection using an infrastructure camera including ground truth vehicle localization

Intersection Monitoring: A dataset for vehicle detection using an infrastructure camera including ground truth vehicle localization. This dataset contains images from a infrastructure camera monitoring an intersection and the ground-truth information for a vehicle crossing it from different directions. The data is aimed to develop and improve image-based vehicle detection algorithms. The dataset includes two different recordings with the same structure. For each of them, the sequence of images is provided in PNG format, along with the ground truth of the vehicle. The ground truth data was captured using a high-precision GNSS receiver installed on the test vehicle, fused with in-vehicle sensors using an Extended Kalman Filter (EKF) . Time considerations The camera and GNSS receiver clocks were synchronized before each test to have the same time base. Each test last about 140 seconds. Image files Image files can be found on the img/ folder within for each test. The camera was configured to record images at 25 fps: <ul> <li>Test 1: 3704 files</li> <li>Test 2: 3426 files</li> </ul> Vehicle localization ground truth The test vehicle is a prototype of Autonomous Vehicle developed by the <a href="https://autopia.car.upm-csic.es">AUTOPIA</a> research group at the <a href="https://car.upm-csic.es">Centre for Automation and Robotics</a> in Spain. Vehicle location mainly depends on a Trimble BX982 GNSS receiver using RTK inputs from a local station. However, the location algorithm applies an EKF for combining GNSS measurements with different onboard sensors providing yaw rate, longitudinal acceleration and speed, steering wheel position and speed, etc. For each test, the vehicle.csv file contains the vehicle information recorded at 20Hz. This file includes: <ul> <li>UTM Time: Time using the format HHMMSSss: <ul> <li>HH: Hours</li> <li>MM: Minutes</li> <li>SS: Seconds</li> <li>ss: Fraction of seconds</li> </ul> </li> <li>UTM East: East coordinate of the GNSS antenna in the UTM frame, in meters.</li> <li>UTMNorth: North coordinate of the GNSS antenna in the UTM frame, in meters.</li> <li>Orientation: Yaw angle of the vehicle, measured from the East axis (x-axis).</li> <li>Speed: Vehicle speed in m/s</li> <li>Acceleration: Vehicle acceleration in m/s^2</li> </ul> Camera info The file camera_parameters.json includes all the information about the intrinsic and extrinsic parameters for the camera. The configuration stored on this file appplies to all tests. The camera used is an AXIS M1125 with variable focal length. It was installed in a communication tower near the intersection. Vehicle info The file vehicle_parameters.json includes information about vehicle dimensions and antenna location. The GNSS antenna is installed near the rear axle of the vehicle, in the middle part of the vehicle. The configuration stored on this file appplies to all tests. Tools Some MATLAB tools can be found in <a href="https://github.com/autopia-car/datasets-tools-intersection-monitoring">https://github.com/autopia-car/datasets-tools-intersection-monitoring</a> Disclaimer That the dataset comes "AS IS", without express or implied warranty and/or any liability exceeding mandatory statutory obligations. This especially applies to any obligations of care or indemnification in connection with the dataset. The dataset was created for our research purposes only and no quality assessment was done for the usage in products of any kind. We can therefore not guarantee for the correctness, completeness or reliability of the provided data set.

opencc-by-4.0Mar 2023View details →

Driving environmental images from on-board camera and RTK-based localization for autonomous vehicles

This dataset contains images from an on-board frontal camera as well as information of the vehicle state, including: accurate global localization, heading, speed, acceleration and steering angle position. The data was captured from one of the vehicle instrumenhted vehicles of the <a href="https://autopia.car.upm-csic.es/">AUTOPIA group</a> at the surrounding of the <a href="https://www.car.upm-csic.es/">Centre for Automation and Robotics</a>, in Arganda del Rey (Madrid, Spain). The dataset is aimed to help users in developing and improving image-based road detection algorithms, as well as machine learning end-to-end approaches using driving information provided (steering angle, vehicle speed, etc.). The dataset has also been used to develop algorithms for online map adaptation based on computer vision. Please, refer to: "Artuñedo, A. (2020). Decision-making Strategies for Automated Driving in Urban Environments. In Springer Theses. Springer International Publishing. <a href="https://doi.org/10.1007/978-3-030-45905-5">https://doi.org/10.1007/978-3-030-45905-5</a>" Image information The sequence of images is provided in PNG format, and is named following the pattern: 'l_sA_nsB.png', where A and B are the second and nanosecond when the image was captured, so that A+B*1e-9 is the capture time. The images were captuted with the camera Bumblebee 2 0,8 MP Color FireWire 1394a 3,8mm (Sony ICX204) with the following technical features: <ul> <li>Sensor type: CCD</li> <li>Sensor format: 1/3"</li> <li>Pixel size 4.65 µm</li> <li>Focal length: 3.8mm</li> <li>Fames per second: 20 fps</li> <li>Resolution: 1024 x 768</li> </ul> Camera position The camera is placed at a height 1290 mm measured from the ground plane. With respect to the vehicle frame, it has a yaw of -2º and a pitch 5.5º, in order to focus the field of view in the road.   Vehicle localization The test vehicle is a prototype of Autonomous Vehicle developed by the <a href="https://autopia.car.upm-csic.es/">AUTOPIA reseach group</a> at the Centre for Automation and Robotics in Spain. Vehicle location mainly depends on a Trimble BX982 GNSS receiver using RTK. The GNSS antenna is installed near the rear axle of the vehicle, in the middle part of the vehicle. However, the location algorithm applies an EKF for combining GNSS measurements with different onboard sensors providing yaw rate, longitudinal acceleration and speed, steering wheel position and speed, etc. The vehicle localization file ('vehicle_data.csv') includes the following data at a sample rate of 20 Hz: <ul> <li>Time (s)</li> <li>UTM East (m)</li> <li>UTM North (m)</li> <li>Orientation (º)</li> <li>Speed (m/s)</li> <li>Acceleration(m/s^2)</li> <li>Steering wheel angle (º)</li> </ul> Disclaimer That the dataset comes "AS IS", without express or implied warranty and/or any liability exceeding mandatory statutory obligations. This especially applies to any obligations of care or indemnification in connection with the dataset. The dataset was created for our research purposes only and no quality assessment was done for the usage in products of any kind. We can therefore not guarantee for the correctness, completeness or reliability of the provided data set.

opencc-zeroApr 2023View details →

Mammalian Camera Trap Data; Northwest Arkansas

The human footprint is rapidly expanding, and wildlife habitat is continuously being converted to human residential properties. Surviving wildlife that reside in developing areas are displaced to nearby undeveloped areas. However, some animals can co-exist with humans and acquire the necessary resources (food, water, shelter) within the human environment. This may be particularly true when development is low intensity, as in residential suburban yards. Yards are individually managed "greenspaces" that can provide a range of food (e.g., bird feeders, compost, gardens), water (bird baths and garden ponds), and shelter resources (e.g., brush-piles, outbuildings) and are surrounded by varying landscape cover. To evaluate which residential landscape and yard features influence the richness and diversity of mammalian herbivores and mesopredators; we deployed wildlife game cameras in 46 residential yards in summer 2021 and 96 yards in summer 2022. We found that mesopredator diversity had a negative relationship with fences and was positively influenced by the number of bird feeders present in a yard. Mesopredator richness increased with the amount of forest within 400m of the camera. Herbivore diversity and richness were positively correlated to the area of forest within 400m surrounding yard and by garden area within yards, respectively. Our results suggest that while landscape does play a role in the presence of wildlife in a residential area, homeowners also have agency over the richness and diversity of mammals occurring in their yards based on the features they create or maintain on their properties.

opencc-by-nc-4.0Apr 2023View details →

Wide-Area Geolocalization with a Limited Field of View Camera in Challenging Urban Environments Dataset

Rectified imagery with latitude and longitude tags for paths driven throughout Cambridge and Boston, MA in November 2022. Faces and license plates have been blurred for anonymity. Dataset created for “Wide-Area Geolocalization with a Limited Field of View Camera in Challenging Urban Environments” paper, currently under review in IEEE Transactions on Robotics. Lena M. Downes is a Draper Scholar at MIT.

opencc-by-4.0Apr 2023View details →

Meteor Project Camera and VLF Receiver Data

Meteor project camera and VLF receiver data used in the Figures 4,5, and 8 of Vankawala et al. 2023. Should be used with the codes found on <a href="https://github.com/parakshv/Meteor_Project">Github</a> in order to reproduce the plots.

opencc-by-4.0Apr 2023View details →

Adapting camera-trap placement based on animal behaviour for rapid detection: a focus on the Endangered, white-bellied pangolin (Phataginus tricuspis)

Table containing detection data of species using two camera trap placement strategies (log vs non-log)

opencc-by-4.0May 2023View details →

CARLA dataset for monocular depth estimation with varying camera parameters

Dataset contains images and corresponding ground truth depths in CARLA simulator. Dataset has been collected within urban, rural, and highway environments across 8 different maps Town01 - Town07 and Town10HD. Each image is created with different camera parameters (camera pitch, camera height and focal length) sampled from uniform distribution.

Topological characterization of the retinal microvascular network visualized by portable fundus camera- effects of chronic disease (TREND2) database

Introduction Topological characterization of the Retinal microvascular nEtwork visualized by portable fuNDus camera (TREND 2) is a database of digital color eye fundus images created as an addition to TREND  database (https://zenodo.org/badge/DOI/10.5281/zenodo.4521044.svg). TREND 2 databse was created by medical professionals of the Faculty of Medicine of the University of Montenegro in 2023.   Purpose 1) to provide a standard that defines normal and abnormal retinal anatomy and microvascular geometry as it appears when visualized by the portable fundus camera 2) to help the development of new methods for stratification of the risk for the development of various eye diseases, as well as systemic diseases that affect microvasculature 3) to aid the development of biomarkers of accelerated aging 4) to provide a standard that can be used to develop software for segmentation of retinal microvasculature, grading the quality of retinal digital images, and computer-aided diagnosis of systemic and chronic diseases. All color digital images were acquired with a hand-held portable, non-mydriatic MiiS HORUS Scope DEC 200 with 45º FOV and 2560 X 1920 pixel resolution.   Data The TREND public database contains 28 color fundus images of old subjects (20 images from subjects with one or more chronic diseases such as type 2 diabetes mellitus, hypertension or Alzheimer's dementia- O_CD group, and 8 images from subjects with no chronic diseases- O_NCD group). Each image is associated with a corresponding binarized image of the manually segmented microvascular network. <table> <caption>Inclusion and Exclusion Criteria</caption> <thead> <tr> <th scope="col">O_NCD group</th> <th scope="col">O_CD group</th> </tr> </thead> <tbody> <tr> <td>Inclusion Criteria</td> <td>Inclusion Criteria</td> </tr> <tr> <td>- at least 56 years old</td> <td>- at least 56 years old</td> </tr> <tr> <td> - no current acute disease - no history of alcohol, or drug abuse, or psychiatric disease </td> <td> - no current acute disease - no history of alcohol, or drug abuse, or psychiatric disease </td> </tr> <tr> <td> - no history of alcohol, or drug abuse, or psychiatric disease </td> <td>- no history of alcohol, or drug abuse, or psychiatric disease</td> </tr> <tr> <td>- negative history of any chronic disease</td> <td> - controlled hypertension (blood pressure<140/90 mmHg), and/or - controlled type 2 diabetes mellitus, and/or - Alzheimer's dementia </td> </tr> <tr> <td>Exclusion Criteria</td> <td>Exclusion Criteria</td> </tr> <tr> <td> - presence of opacities of the transparent media in both eyes affecting - myopia ≥5 diopters </td> <td> - presence of opacities of the transparent media in both eyes affecting - myopia ≥5 diopters </td> </tr> </tbody> </table> Files: 1_OLD WITH CHRONIC DISEASE_RAW (20 images in tif format) 2_OLD WITH CHRONIC DISEASE_SEGMENTED (20 images in png format) 3_OLD WITH NO CHRONIC DISEASE_RAW (8 images in tif format) 4_OLD WITH NO CHRONIC DISEASE SEGMENTED (8 images in png format) 5_ASSOCIATED DATA (xslx format) 6_RETINAL PATHOLOGY (docx format)

opencc-zeroMay 2023View details →

Camera trap grey squirrel photograph data

Effective wildlife population management requires an understanding of the abundance of the target species. In the UK, the increase in numbers and range of the non-native invasive grey squirrel Sciurus carolinensis poses a substantial threat to the existence of the native red squirrel S. vulgaris, to tree health, and to the forestry industry. Reducing the number of grey squirrels is crucial to mitigate their impacts. Camera traps are increasingly used to estimate animal abundance, and methods have been developed that do not require the identification of individual animals. Most of these methods have been focussed on medium to large mammal species with large range sizes and may be unsuitable for measuring local abundances of smaller mammals that have variable detection rates and hard-to-measure movement behaviour. The aim of this study was to develop a practical and cost-effective method, based on a camera trap index, that could be used by practitioners to estimate target densities of grey squirrels in woodlands to provide guidance on the numbers of traps or contraceptive feeders required for local grey squirrel control. Camera traps were deployed in ten independent woods of between 6 and 28 ha in size. An index, calculated from the number of grey squirrel photographs recorded per camera per day had a strong linear relationship (R2 = 0.90) with the densities of squirrels removed in trap and dispatch operations. From different time filters tested, a 5 minute filter was applied, where photographs of squirrels recorded on the same camera within 5 minutes of a previous photograph were not counted. There were no significant differences between the number of squirrel photographs per camera recorded by three different models of camera, increasing the method's practical application. This study demonstrated that a camera index could be used to inform the number of feeders or traps required for grey squirrel management through culling or contraception. Results could be obtained within six days without requiring expensive equipment or a high level of technical input. This method can easily be adapted to other rodent or small mammal species, making it widely applicable to other wildlife management interventions.

opencc-by-4.0Jun 2023View details →

ENDGAME - Laboratory Experiment 2022-12-01 Exp. 001 - Part 3 - High Speed Camera data

Preliminary test with high speed camera and Schlieren shadow photography. Images of the injection of air bubbles in a 2D setup obtained using 2 parallel Plexiglas sheets (10 mm thickness) separated by rubber seals and filled with a distilled water. The gap between the two parallel sheets is 3 mm. Air was injected manually into the 2D setup through a capillary tube (~2 mm diam). Frame rate of the high speed camera is 250 fps. The spherical mirror used for the Schlieren setup was 75 mm wide with a 750 mm focal length.

opencc-zeroJul 2023View details →

Estimating animal location from non-overhead camera views

Tracking an animal's location from video has many applications, from providing information on health and welfare to validating sensor-based technologies. Typically, accurate location estimation from video is achieved using cameras with overhead (top-down) views, but structural and financial limitations may require mounting cameras at other angles. We describe a user-friendly solution to manually extract an animal's location from non-overhead video. Our method uses QGIS, an open-source geographic information system, to: (1) assign facility-based coordinates to pixel coordinates in non-overhead frames; 2) use the referenced coordinates to transform the non-overhead frames to an overhead view; and 3) determine facility-based x, y coordinates of animals from the transformed frames. Using this method, we could determine an object's facility-based x, y coordinates with an accuracy of 0.13 ± 0.09 m (mean ± SD; range: 0.01–0.47 m) when compared to the ground truth (coordinates manually recorded with a laser tape measurer). We demonstrate how this method can be used to answer research questions about space-use behaviors in captive animals, using 6 ewe-lamb pairs housed in a group pen. As predicted, we found that lambs maintained closer proximity to their dam compared to other ewes in the group and lamb-dam range sizes were strongly correlated. However, the distance traveled by lambs and their dams did not correlate, suggesting that activity levels differed within the pair. This method demonstrates how user-friendly, open-source GIS tools can be used to accurately estimate animal location and derive space-use behaviors from non-overhead video frames. This method will expand capacity to obtain spatial data from animals in facilities where it is not possible to mount cameras overhead.