Understanding External Influences on Target Detection and Classification Using Camera Trap Images and Machine Learning

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT REQUEST OF AUTHOR.] Camera traps are a popular tool to sample animal populations because they are noninvasive, detect a variety of species, and can record many thousands of animal detections per deployment. Cameras are typically set to take bursts of multiple images for each detection, and are deployed in arrays of dozens or hundreds of sites, often resulting in millions of images per study. The task of converting images to animal detection records from such large image collections is daunting, and made worse by situations that generate copious empty pictures from false triggers (e.g. camera malfunction or moving vegetation) or pictures of humans. We offer the first widely available computer vision tool for processing camera trap images. Our results show that the tool is accurate and results in substantial time savings for processing large image datasets, thus improving our ability to monitor wildlife across large scales with camera traps. In this dissertation, we have developed new image/video processing and computer vision algorithms for efficient and accurate object detection and sequence-level classiffication from natural scene camera-trap images. This work addresses the following five major tasks: (1) Human-animal detection. We develop a fast and accurate scheme for human-animal detection from highly cluttered camera-trap images using joint background modeling and deep learning classification. Specifically, first, We develop an effective background modeling and subtraction scheme to generate region proposals for the foreground objects. We then develop a cross-frame image patch verification to reduce the number of foreground object proposals. Finally, We perform complexity-accuracy analysis of deep convolutional neural networks (DCNN) to develop a fast deep learning classification scheme to classify these region proposals into three categories: human, animals, and background patches. The optimized DCNN is able to maintain high level of accuracy while reducing the computational complexity by 14 times. Our experimental results demonstrate that the proposed method outperforms existing methods on the camera-trap dataset. (2) Object segmentation from natural scene. We first design and train a fast DCNN for animal-human-background object classification, which is used to analyze the input image to generate multi-layer feature maps, representing the responses of different image regions to the animal-human-background classifier. From these feature maps, we construct the so-called deep objectness graph for accurate animal-human object segmentation with graph cut. The segmented object regions from each image in the sequence are then verfied and fused in the temporal domain using background modeling. Our experimental results demonstrate that our proposed method outperforms existing state-of-the-art methods on the camera-trap dataset with highly cluttered natural scenes. (3) DCNN domain background modeling. We replaced the background model with a new more efficient deep learning based model. The input frames are segmented into regions through the deep objectness graph then the region boundaries of the input frames are multiplied by each other to obtain the regions of movement patches. We construct the background representation using the temporal information of the co-located patches. We propose to fuse the subtraction and foreground/background pixel classiffcation of two representation : a) chromaticity and b) deep pixel information. (4) Sequence-level object classiffcation. We proposed a new method for sequence-level video recognition with application to animal species recognition from camera trap images. First, using background modeling and cross-frame patch verification, we developed a scheme to generate candidate object regions or object proposals in the spatiotemporal domain. Second, we develop a dynamic programming optimization approach to identify the best temporal subset of object proposals. Third, we aggregate and fuse the features of these selected object proposals for efficient sequence-level animal species classification.

Most machine learning methods for animal recognition in camera trap images are limited to mammal identification and group birds into a single class. Machine learning methods for visually discriminating birds, in turn, cannot discriminate between mammals and are not designed for camera trap images. The authors present deep neural network models to recognise both mammals and bird species in camera trap images. They train neural network models for species classification as well as for predicting the animal taxonomy, that is, genus, family, order, group, and class names. Different neural network architectures, including ResNet, EfficientNetV2, Vision Transformer, Swin Transformer, and ConvNeXt, are compared for these tasks. Furthermore, the authors investigate approaches to overcome various challenges associated with camera trap image analysis. The authors’ best species classification models achieve a mean average precision (mAP) of 97.91% on a validation data set and mAPs of 90.39% and 82.77% on test data sets recorded in forests in Germany and Poland, respectively. Their best taxonomic classification models reach a validation mAP of 97.18% and mAPs of 94.23% and 79.92% on the two test data sets, respectively.

Automated monitoring technologies can increase the efficiency of ecological data collection and support data‐driven conservation. Camera traps coupled with infrared light barriers can be used to monitor temperate‐zone bat assemblages at underground hibernacula, where thousands of individuals of multiple species can aggregate in winter. However, the broad‐scale adoption of such photo‐monitoring techniques is limited by the time‐consuming bottleneck of manual image processing. Here, we present BatNet, an open‐source, deep learning‐based tool for automated identification of 13 European bat species from camera trap images. BatNet includes a user‐friendly graphical interface, where it can be retrained to identify new bat species or to create site‐specific models to improve detection accuracy at new sites. Model accuracy was evaluated on images from both trained and untrained sites, and in an ecological context, where community‐ and species‐level metrics (species diversity, relative abundance, and species‐level activity patterns) were compared between human experts and BatNet. At trained sites, model performance was high across all species (F1‐score: 0.98–1). At untrained sites, overall classification accuracy remained high (96.7–98.2%), when camera placement was comparable to the training images (<3 m from the entrance; <45° angle relative to the opening). For atypical camera placements (>3 m or >45° angle), retraining the detector model with 500 site‐specific annotations achieved an accuracy of over 95% at all sites. In the ecological case study, all investigated metrics were nearly identical between human experts and BatNet. Finally, we exemplify the ability to retrain BatNet to identify a new bat species, achieving an F1‐score of 0.99 while maintaining high classification accuracy for all original species. BatNet can be implemented directly to scale up the deployment of camera traps in Europe and enhance bat population monitoring. Moreover, the pretrained model can serve as a baseline for transfer learning to automatize the image‐based identification of bat species worldwide.

Camera traps are widely used to monitor the biodiversity and population density of animal species. Camera trap images are usually taken in bursts, and the animal counting problem for a sequence of camera trap images is also an important part of evaluating animal population density. In this paper, two new animal counting methods based on Microsoft MegaDetector V 4 have been proposed. FilterDetector uses different filters with bounding box ensemble algorithms to achieve more accurate bounding box detection. DLEDetector is an ensemble method that uses two base deep learning models to correct and enhance the detection result of MegaDetector. Our experimental results in iWildCam 2022 competition test dataset show that both methods outperformed the best method in iWildCam 2021 and the baseline method based on MegaDetector V 4 in iWildCam 2022 competition by 9.09% and 6.44%, respectively, and ranked first and third in the competition.

As camera trapping has become a standard practice in wildlife ecology, developing techniques to extract additional information from images will increase the utility of generated data. Despite rapid advancements in camera trapping practices, methods for estimating animal size or distance from the camera using captured images have not been standardized. Deriving animal sizes directly from images creates opportunities to collect wildlife metrics such as growth rates or changes in body condition. Distances to animals may be used to quantify important aspects of sampling design such as the effective area sampled or distribution of animals in the camera's field‐of‐view. We present a method of using pixel measurements in an image to estimate animal size or distance from the camera using a conceptual model in photogrammetry known as the ‘pinhole camera model’. We evaluated the performance of this approach both using stationary three‐dimensional animal targets and in a field setting using live captive reindeer Rangifer tarandus ranging in size and distance from the camera. We found total mean relative error of estimated animal sizes or distances from the cameras in our simulation was −3.0% and 3.3% and in our field setting was −8.6% and 10.5%, respectively. In our simulation, mean relative error of size or distance estimates were not statistically different between image settings within camera models, between camera models or between the measured dimension used in calculations. We provide recommendations for applying the pinhole camera model in a wildlife camera trapping context. Our approach of using the pinhole camera model to estimate animal size or distance from the camera produced robust estimates using a single image while remaining easy to implement and generalizable to different camera trap models and installations, thus enhancing its utility for a variety of camera trap applications and expanding opportunities to use camera trap images in novel ways.

AnimalFinder: A semi-automated system for animal detection in time-lapse camera trap images

Feral cats are a key threat to many Australian native fauna, with camera traps increasingly used to identify individuals for evaluation of management actions. However, observer bias and camera trap settings can affect individual identification rates. We compared feral cat individual identification by two observers with extremes of experience. Arrays of 39–50 camera traps were deployed continuously for 22 months at four sites in the Western Division of New South Wales. Where possible, feral cats were individually identified from phenotypic characteristics by an expert and naïve lay observer. We obtained 10 465 feral cat images, with 72 cats individually identified across the sites. The experienced observer attributed more feral cat events to a known individual compared with the lay observer (21.3 vs 12.9%, respectively). Forty three percent of cat images were similarly tagged by both observers. Daytime events yielded higher identification rates and match success (28.1 vs 19.5 and 17.9 vs 11.8% for day vs night events for the expert and lay observer, respectively). Lack of congruence between observers, combined with a small number of events where cats could be individually identified, and differences in identification accuracy over time and between sites, makes estimation of detection probabilities and errors difficult.

Birds are important indicators for monitoring both biodiversity and habitat health; they also play a crucial role in ecosystem management. Declines in bird populations can result in reduced ecosystem services, including seed dispersal, pollination and pest control. Accurate and long-term monitoring of birds to identify species of concern while measuring the success of conservation interventions is essential for ecologists. However, monitoring is time-consuming, costly and often difficult to manage over long durations and at meaningfully large spatial scales. Technology such as camera traps, acoustic monitors and drones provide methods for non-invasive monitoring. There are two main problems with using camera traps for monitoring: (a) cameras generate many images, making it difficult to process and analyse the data in a timely manner; and (b) the high proportion of false positives hinders the processing and analysis for reporting. In this paper, we outline an approach for overcoming these issues by utilising deep learning for real-time classification of bird species and automated removal of false positives in camera trap data. Images are classified in real-time using a Faster-RCNN architecture. Images are transmitted over 3/4G cameras and processed using Graphical Processing Units (GPUs) to provide conservationists with key detection metrics, thereby removing the requirement for manual observations. Our models achieved an average sensitivity of 88.79%, a specificity of 98.16% and accuracy of 96.71%. This demonstrates the effectiveness of using deep learning for automatic bird monitoring.

ABSTRACTThe Grévy's zebra, an endangered species native to Kenya and southern Ethiopia, has been the target of sustained conservation efforts in recent years. Accurately monitoring Grévy's zebra populations is essential for ecologists to evaluate ongoing conservation initiatives. Recently, in both 2016 and 2018, a full census of the Grévy's zebra population was enabled by the Great Grévy's Rally (GGR), a citizen science event that combines teams of volunteers to capture data with computer vision algorithms that help experts estimate the number of individuals in the population. A complementary, scalable, cost‐effective and long‐term Grévy's population monitoring approach involves deploying a network of camera traps, which we have done at the Mpala Research Centre in Laikipia County, Kenya. In both scenarios, a substantial majority of the images of zebras are not usable for individual identification due to ‘in‐the‐wild’ imaging conditions—occlusions from vegetation or other animals, oblique views, low image quality and animals that appear in the far background and are thus too small to identify. Camera trap images, without an intelligent human photographer to select the framing and focus on the animals of interest, are of even poorer quality, with high rates of occlusion and high spatiotemporal similarity within image bursts. We employ an image filtering pipeline incorporating animal detection, species identification, viewpoint estimation, quality evaluation and temporal subsampling to compensate for these factors and obtain individual crops from camera trap and GGR images of suitable quality for re‐ID. We then employ the local clusterings and their alternatives (LCA) algorithm, a hybrid computer vision and graph clustering method for animal re‐ID, on the resulting high‐quality crops. Our method processed images taken during GGR‐16 and GGR‐18 in Meru County, Kenya, into 4142 highly comparable annotations, requiring only 120 contrastive same‐vs‐different‐individual decisions from a human reviewer to produce a population estimate of 349 individuals (within 4.6 of the ground truth count in Meru County). Our method also efficiently processed 8.9M unlabelled camera trap images from 70 camera traps at Mpala over 2 years into 685 encounters of 173 unique individuals, requiring only 331 contrastive decisions from a human reviewer.

Camera trapping is widely used to monitor mammalian wildlife but creates large image datasets that must be classified. In response, there is a trend towards crowdsourcing image classification. For high‐profile studies of charismatic faunas, many classifications can be obtained per image, enabling consensus assessments of the image contents. For more local‐scale or less charismatic communities, however, demand may outstrip the supply of crowdsourced classifications. Here, we consider MammalWeb, a local‐scale project in North East England, which involves citizen scientists in both the capture and classification of sequences of camera trap images. We show that, for our global pool of image sequences, the probability of correct classification exceeds 99% with about nine concordant crowdsourced classifications per sequence. However, there is high variation among species. For highly recognizable species, species‐specific consensus algorithms could be even more efficient; for difficult to spot or easily confused taxa, expert classifications might be preferable. We show that two types of incorrect classifications – misidentification of species and overlooking the presence of animals – have different impacts on the confidence of consensus classifications, depending on the true species pictured. Our results have implications for data capture and classification in increasingly numerous, local‐scale citizen science projects. The species‐specific nature of our findings suggests that the performance of crowdsourcing projects is likely to be highly sensitive to the local fauna and context. The generality of consensus algorithms will, thus, be an important consideration for ecologists interested in harnessing the power of the crowd to assist with camera trapping studies.

Understanding the spatial and temporal distribution of hunting is essential for the management of protected areas, as well as for assessing its impacts on wildlife. This knowledge also allows patrols to be directed toward areas under greater hunting pressure. A promising approach to estimating hunting pressure is through the use of autonomous bioacoustic recorders combined with artificial intelligence tools. However, it is necessary to develop a process capable of accurately describing the spatial and temporal distribution of hunting and, in parallel with wildlife monitoring, understanding its ecological impacts. Therefore, our goal is to estimate hunting pressure across space and time and to assess its effects on forest ungulate species. We intend to evaluate the impact of hunting by testing its influence on the density, occupancy, and activity patterns of ungulates. To achieve this, we established a 1,500-ha sampling grid in Iguaçu National Park, where each of the 24 plots is equipped with one bioacoustic recorder and three camera traps, which will operate for two years. So far, we have completed four months of sampling. We also conducted a preliminary experiment to record gunshots from different types of firearms and ammunition at various distances from the recorders, aiming to model sound intensity as a function of distance. As a preliminary result, we found that gunshots from 12-gauge firearms can be detected with an intensity of 20 dB at distances of up to 900 meters, indicating this as an optimal spacing between recorders for triangulation of detected gunshots. uring the setup of the sampling grid, we already found seven clear pieces of evidence of poaching activity in different locations within the study area, including platforms used for poaching (2), campsites (2), artificial salt licks (1), and cartridges (2), and we recorded a poacher with a camera trap. In the coming months, hunting activity will be modeled spatially and temporally. The resulting database of bioacoustic recordings and camera-trap images will be made available to the scientific community to foster future collaborations and improve conservation strategies. Funded by FAPESP #2024/18433-5

<p><span lang="AF">The foundation of systematic research depends on precise species identification, functioning as a critical component in the processes of biological research. Wildlife biologists are prompting for more effective techniques to fulfill the expanding need for species identification. The rise in open source image data showing animal species, captured by digital cameras and other digital methods of collecting data, has been monumental. This rapid expansion of animal image data, integrated with state-of-the-art machine learning techniques such as deep learning which has shown significant capabilities for automating species identification. This paper focuses on the role of deep neural network architectures in furthering technological advancements in automating species identification in recent years. To advocate further investigation in this field, an examination of machine learning architectures for species identification was presented in this work. This examination focuses primarily on image analyses and discusses their significance in wildlife conservation. Fundamentally, the aim of this article is to offer insights into the present advancements in automating species identification and to act as a reference for scholars who are keen to integrate machine learning techniques into ecological studies. Systems designed through Artificial Intelligence are extensive in providing toolkits for systematic identification of species in the upcoming years.</span></p>

Accurately estimating population sizes for free-ranging animals through noninvasive methods, such as camera trap images, remains particularly limited by small datasets. To overcome this, we developed a flexible model for estimating upper limit populations and exemplified it by studying a group-living synanthrope, the long-tailed macaque (Macaca fascicularis). Habitat preference maps, based on environmental and GPS data, were generated with a maximum entropy model and combined with data obtained from camera traps, line transect distance sampling, and direct sightings to produce an expected number of individuals. The mapping between habitat preference and number of individuals was optimized through a tunable parameter ρ (inquisitiveness) that accounts for repeated observations of individuals. Benchmarking against published data highlights the high accuracy of the model. Overall, this approach combines citizen science with scientific observations and reveals the long-tailed macaque populations to be (up to 80%) smaller than expected. The model's flexibility makes it suitable for many species, providing a scalable, noninvasive tool for wildlife conservation.

This paper presents the evaluation of 36 convolutional neural network (CNN) models, which were trained on the same dataset (ImageNet). The aim of this research was to evaluate the performance of pre-trained models on the binary classification of images in a “real-world” application. The classification of wildlife images was the use case, in particular, those of the Eurasian lynx (lat. “Lynx lynx”), which were collected by camera traps in various locations in Croatia. The collected images varied greatly in terms of image quality, while the dataset itself was highly imbalanced in terms of the percentage of images that depicted lynxes.

Optimized machine learning approaches for identifying vertical temperature gradient on ballastless track in natural environments