O’Neill Automated species identification: why not

Bioacoustics, 2007

... TODOR GANCHEV† AND ILYAS POTAMITIS‡ †Department of Electrical and Computer Engineering, University of Patras, 26500 Rion-Patras, Greece E-mail: tganchev@wcl.ee.upatras. gr ‡Department of Music Technology and Acoustics, Technological ... 1957; Shaw 1975). ...

Journal of Insect Conservation, 2014

Acoustic emissions of animals serve communicative purposes and most often contain species-specific and individual information exploitable to listeners, rendering bioacoustics a valuable tool for biodiversity monitoring.

Systematic Entomology, 2010

Species in cryptic complexes tend to be very difficult, if not impossible, to identify using morphological characters. One such complex is the spruce budworm (Choristoneura fumiferana Clemens, 1865) species group, an economically important group of Nearctic forest pests. Morphological, ecological, behavioural and genetic characters have been studied to try to understand the taxonomy of this group better, but diagnostic character states differ in frequency rather than being complete replacements between each species. We used mitochondrial DNA (mtDNA), together with a new morphology-based character system that focuses on forewing colour components, to determine if one or a combination of character sources can be used for species diagnoses within the spruce budworm complex. We characterized 47 forewing morphometric measurements and sequenced a 470 bp region of cytochrome c oxidase I mtDNA for 111 ingroup individuals comprising five taxa within the complex. Larval host association and coloration or adult pheromone attraction were used as the prior method for grouping individuals. Our results showed that linear discriminant analysis of morphometric wing characters gave unique clusters for all species on the first and second canonical axes, except for a partial overlap between C. fumiferana and C. biennis, which are not sympatric in nature. In contrast, mtDNA distinguished C. fumiferana, C. pinus pinus Freeman, 1953 and a group of western species, but the three western species (C. occidentalis Freeman, 1967, C. biennis Freeman, 1967 and C. lambertiana Busck, 1915 shared mtDNA haplotypes. On the basis of the linear discriminant analysis of the combined character set, this study supports the application of both morphology and mtDNA within a framework of integrative taxonomy as the most accurate method for species identification. Furthermore, it demonstrates the utility of quantitative colour analysis, which may be particularly helpful for groups in which colour characters are difficult to divide into discrete units due to intergrading hues.

Lecture Notes in Computer Science, 2011

In environmental surveillance, ecology experts use a stan-10 dard tracking tunnel system to acquire tracks or footprints of small 11 animals, so that they can measure the presence of any selected ani-12 mals or detect threatened species based on the manual analysis of gath-13 ered tracks. Unfortunately, distinguishing morphologically similar species 14 through analysing their footprints is extremely difficult, and even very 15 experienced experts find it hard to provide reliable results on footprint 16 identification. This expensive task also requires a great amount of efforts 17 on observation. In recent years, image processing technology has become 18 a model example for applying computer science technology to many other 19 study areas or industries, in order to improve accuracy, productivity, and 20 reliability. In this paper, we propose a method based on image processing 21 technology, it firstly detects significant interest points from input track-22 ing card images. Secondly, it filters irrelevant interest points in order 23 to extract regions of interest. Thirdly, it gathers useful information of 24 footprint geometric features, such as angles, areas, distance, and so on.

BioEssays, 2011

DNA barcodes, like traditional sources of taxonomic information, are potentially powerful heuristics in the identification of described species but require mindful analytical interpretation. The role of DNA barcoding in generating hypotheses of new taxa in need of formal taxonomic treatment is discussed, and it is emphasized that the recursive process of character evaluation is both necessary and best served by understanding the empirical mechanics of the discovery process. These undertakings carry enormous ramifications not only for the translation of DNA sequence data into taxonomic information but also for our comprehension of the magnitude of species diversity and its disappearance. This paper examines the potential strengths and pitfalls of integrating DNA sequence data, specifically in the form of DNA barcodes as they are currently generated and analyzed, with taxonomic practice.

PLoS ONE, 2013

Background: The explosively radiating evolution of cichlid fishes of Lake Malawi has yielded an amazing number of haplochromine species estimated as many as 500 to 800 with a surprising degree of diversity not only in color and stripe pattern but also in the shape of jaw and body among them. As these morphological diversities have been a central subject of adaptive speciation and taxonomic classification, such high diversity could serve as a foundation for automation of species identification of cichlids.

Ecology and Evolution, 2015

Microbes are critical components of ecosystems and provide vital services (e.g., photosynthesis, decomposition, nutrient recycling). From the diverse roles microbes play in natural ecosystems, high levels of functional diversity result. Quantifying this diversity is challenging, because it is weakly associated with morphological differentiation. In addition, the small size of microbes hinders morphological and behavioral measurements at the individual level, as well as interactions between individuals. Advances in microbial community genetics and genomics, flow cytometry and digital analysis of still images are promising approaches. They miss out, however, on a very important aspect of populations and communities: the behavior of individuals. Video analysis complements these methods by providing in addition to abundance and trait measurements, detailed behavioral information, capturing dynamic processes such as movement, and hence has the potential to describe the interactions between individuals. We introduce BEMOVI, a package using the R and ImageJ software, to extract abundance, morphology, and movement data for tens to thousands of individuals in a video. Through a set of functions BEMOVI identifies individuals present in a video, reconstructs their movement trajectories through space and time, and merges this information into a single database. BEMOVI is a modular set of functions, which can be customized to allow for peculiarities of the videos to be analyzed, in terms of organisms features (e.g., morphology or movement) and how they can be distinguished from the background. We illustrate the validity and accuracy of the method with an example on experimental multispecies communities of aquatic protists. We show high correspondence between manual and automatic counts and illustrate how simultaneous time series of abundance, morphology, and behavior are obtained from BEMOVI. We further demonstrate how the trait data can be used with machine learning to automatically classify individuals into species and that information on movement behavior improves the predictive ability.

Applied Sciences

This work introduces a new approach for automatic identification of crickets, katydids and cicadas analyzing their acoustic signals. We propose the building of a tool to identify this biodiversity. The study proposes a sound parameterization technique designed specifically for identification and classification of acoustic signals of insects using Mel Frequency Cepstral Coefficients (MFCC) and Linear Frequency Cepstral Coefficients (LFCC). These two sets of coefficients are evaluated individually as has been done in previous studies and have been compared with the fusion proposed in this work, showing an outstanding increase in identification and classification at species level reaching a success rate of 98.07% on 343 insect species.

Fisheries, 2013

The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden.

PeerJ, 2015

Species identification-of importance for most biological disciplines-is not always straightforward as cryptic species hamper traditional identification. Fibre-optic near-infrared spectroscopy (NIRS) is a rapid and inexpensive method of use in various applications, including the identification of species. Despite its efficiency, NIRS has never been tested on a group of more than two cryptic species, and a working routine is still missing. Hence, we tested if the four morphologically highly similar, but genetically distinct ant species Tetramorium alpestre, T. caespitum, T. impurum, and T. sp. B, all four co-occurring above 1,300 m above sea level in the Alps, can be identified unambiguously using NIRS. Furthermore, we evaluated which of our implementations of the three analysis approaches, partial least squares regression (PLS), artificial neural networks (ANN), and random forests (RF), is most efficient in species identification with our data set. We opted for a 100% classification certainty, i.e., a residual risk of misidentification of zero within the available data, at the cost of excluding specimens from identification. Additionally, we examined which strategy among our implementations, one-vs-all, i.e., one species compared with the pooled set of the remaining species, or binary-decision strategies, worked best with our data to reduce a multi-class system to a two-class system, as is necessary for PLS. Our NIRS identification routine, based on a 100% identification certainty, was successful with up to 66.7% of unambiguously identified specimens of a species. In detail, PLS scored best over all species (36.7% of specimens), while RF was much less effective (10.0%) and ANN failed completely (0.0%) with our data and our implementations of the analyses. Moreover, we showed that the one-vs-all strategy is the only acceptable option to reduce multi-class systems because of a minimum expenditure of time. We emphasise our classification routine using fibre-optic NIRS in combination with PLS and the one-vs-all strategy as a highly efficient pre-screening identification method for cryptic ant species and possibly beyond.

PeerJ, 2017

Identification of arthropods important in disease transmission is a crucial, yet difficult, task that can demand considerable training and experience. An important case in point is that of the 150+ species of Triatominae, vectors of Trypanosoma cruzi, causative agent of Chagas disease across the Americas. We present a fully automated system that is able to identify triatomine bugs from Mexico and Brazil with an accuracy consistently above 80%, and with considerable potential for further improvement. The system processes digital photographs from a photo apparatus into landmarks, and uses ratios of measurements among those landmarks, as well as (in a preliminary exploration) two measurements that approximate aspects of coloration, as the basis for classification. This project has thus produced a working prototype that achieves reasonably robust correct identification rates, although many more developments can and will be added, and-more broadly-the project illustrates the value of mul...

PeerJ, 2016

Background.Fish species may be identified based on their unique otolith shape or contour. Several pattern recognition methods have been proposed to classify fish species through morphological features of the otolith contours. However, there has been no fully-automated species identification model with the accuracy higher than 80%. The purpose of the current study is to develop a fully-automated model, based on the otolith contours, to identify the fish species with the high classification accuracy.Methods.Images of the right sagittal otoliths of 14 fish species from three families namely Sciaenidae, Ariidae, and Engraulidae were used to develop the proposed identification model. Short-time Fourier transform (STFT) was used, for the first time in the area of otolith shape analysis, to extract important features of the otolith contours. Discriminant Analysis (DA), as a classification technique, was used to train and test the model based on the extracted features.Results.Performance of...

Methods in Ecology and Evolution, 2017

1.Recent technological development has increased our capacity to study the deep sea and the marine benthic realm, particularly with the development of multidisciplinary seafloor observatories. Since 2006, Ocean Networks Canada cabled observatories, have acquired nearly 65 TB and over 90,000 hours of video data from seafloor cameras and Remotely Operated Vehicles (ROVs). Manual processing of these data is time-consuming and highly labour-intensive, and cannot be comprehensively undertaken by individual researchers. These videos are a crucial source of information for assessing natural variability and ecosystem responses to increasing human activity in the deep sea. 2.We compared the performance of three groups of humans and one computer vision algorithm in counting individuals of the commercially important sablefish (or black cod) Anoplopoma fimbria, in recorded video from a cabled camera platform at 900 m depth in a submarine canyon in the Northeast Pacific. The first group of human observers were untrained volunteers recruited via a crowdsourcing platform and the second were experienced university students, who performed the task for their ichthyology class. Results were validated against counts obtained from a scientific expert. 3.All groups produced relatively accurate results in comparison to the expert and all succeeded in detecting patterns and periodicities in fish abundance data. Trained volunteers displayed the highest accuracy and the algorithm the lowest. Please note that this is an author-produced PDF of an article accepted for publication following peer review. The definitive publisher-authenticated version is available on the publisher Web site. 4.As seafloor observatories increase in number around the world, this study demonstrates the value of a hybrid combination of crowdsourcing and computer vision techniques as a tool to help process large volumes of imagery to support basic research and environmental monitoring. Reciprocally, by engaging large numbers of online participants in deep-sea research, this approach can contribute significantly to ocean literacy and informed citizen input to policy development.

Multimedia Tools and Applications, 2015

This paper reports a large-scale experiment aimed at evaluating how state-of-art computer vision systems perform in identifying plants compared to human expertise. A subset of the evaluation dataset used within LifeCLEF 2014 plant identification challenge was therefore shared with volunteers of diverse expertise, ranging from the leading experts of the targeted flora to inexperienced test subjects. In total, 16 human runs were collected and evaluated comparatively to the 27 machine-based runs of LifeCLEF challenge. One of the main outcomes of the experiment is that machines are still far from outperforming the best expert botanists at the image-based plant identification competition. On the other side, the best machine runs are competing with experienced botanists and clearly outperform beginners and inexperienced test subjects. This shows that the performances of automated plant identification

Biodiversity data journal, 2017

Natural history collections contain data that are critical for many scientific endeavors. Recent efforts in mass digitization are generating large datasets from these collections that can provide unprecedented insight. Here, we present examples of how deep convolutional neural networks can be applied in analyses of imaged herbarium specimens. We first demonstrate that a convolutional neural network can detect mercury-stained specimens across a collection with 90% accuracy. We then show that such a network can correctly distinguish two morphologically similar plant families 96% of the time. Discarding the most challenging specimen images increases accuracy to 94% and 99%, respectively. These results highlight the importance of mass digitization and deep learning approaches and reveal how they can together deliver powerful new investigative tools.

BMC Biology

Background Studies of mammalian sexual dimorphism have traditionally involved the measurement of selected dimensions of particular skeletal elements and use of single data-analysis procedures. Consequently, such studies have been limited by a variety of both practical and conceptual constraints. To compare and contrast what might be gained from a more exploratory, multifactorial approach to the quantitative assessment of form-variation, images of a small sample of modern Israeli gray wolf (Canis lupus) crania were analyzed via elliptical Fourier analysis of cranial outlines, a Naïve Bayes machine-learning approach to the analysis of these same outline data, and a deep-learning analysis of whole images in which all aspects of these cranial morphologies were represented. The statistical significance and stability of each discriminant result were tested using bootstrap and jackknife procedures. Results Our results reveal no evidence for statistically significant sexual size dimorphism,...

PLOS ONE

An ongoing study of the genus Merodon Meigen, 1803 in the Republic of South Africa (RSA) has revealed the existence of new species related to M. melanocerus Bezzi, 1915. The M. melanocerus subgroup belongs to the Afrotropical lineage of the M. desuturinus group. Revision of all available material from museums and detailed analyses of newly-collected specimens from our own expeditions to RSA resulted in delimitation of five species: M. capensis Hurkmans sp. n., M. commutabilis Radenković et Vujić sp. n., M. drakonis Vujić et Radenković sp. n., M. flavocerus Hurkmans sp. n. and M. melanocerus. In addition to classical morphological characters, sequences of the mitochondrial COI gene are provided for four related taxa. Results of molecular phylogenetic analyses supports monophyly of the M. desuturinus group and confirmed delimitation between species. Links between Palaearctic and Afrotropical faunas of this group, as well as possible evolutionary paths, are discussed. Based on phylogenetic analyses, four lineages (putative subgenera) have been recognized within the genus Merodon; besides the three previously established ones, albifrons+desuturinus, aureus (sensu lato) and avidus-nigritarsis, one new lineage named natans is distinguished.

Advances in Information Retrieval, 2019

Building accurate knowledge of the identity, the geographic distribution and the evolution of living species is essential for a sustainable development of humanity, as well as for biodiversity conservation. However, the burden of the routine identification of plants and animals in the field is strongly penalizing the aggregation of new data and knowledge. Identifying and naming living plants or animals is actually almost impossible for the general public and often a difficult task for professionals and naturalists. Bridging this gap is a key challenge towards enabling effective biodiversity information retrieval systems. The LifeCLEF evaluation campaign, presented in this paper, aims at boosting and evaluating the advances in this domain since 2011. In particular, the 2019 edition proposes three data-oriented challenges related to the identification and prediction of biodiversity: (i) an image-based plant identification challenge, (ii) a bird sounds identification challenge and (iii) a location-based species prediction challenge based on spatial occurrence data and environmental tensors.

Experimental IR Meets Multilinguality, Multimodality, and Interaction, 2018

Building accurate knowledge of the identity, the geographic distribution and the evolution of living species is essential for a sustainable development of humanity, as well as for biodiversity conservation. Unfortunately, such basic information is often only partially available for professional stakeholders, teachers, scientists and citizens, and often incomplete for ecosystems that possess the highest diversity. In this context, an ultimate ambition is to set up innovative information systems relying on the automated identification and understanding of living organisms as a means to engage massive crowds of observers and boost the production of biodiversity and agro-biodiversity data. The LifeCLEF 2018 initiative proposes three data-oriented challenges related to this vision, in the continuity of the previous editions, but with several consistent novelties intended to push the boundaries of the state-of-the-art in several research directions. This paper describes the methodology of the conducted evaluations as well as the synthesis of the main results and lessons learned.

Experimental IR Meets Multilinguality, Multimodality, and Interaction, 2019

HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

2021

Building accurate knowledge of the identity, the geographic distribution and the evolution of species is essential for the sustainable development of humanity, as well as for biodiversity conservation. However, the difficulty of identifying plants and animals in the field is hindering the aggregation of new data and knowledge. Identifying and naming living plants or animals is almost impossible for the general public and is often difficult even for professionals and naturalists. Bridging this gap is a key step towards enabling effective biodiversity monitoring systems. The LifeCLEF campaign, presented in this paper, has been promoting and evaluating advances in this domain since 2011. The 2021 edition proposes four data-oriented challenges related to the identification and prediction of biodiversity: (i) PlantCLEF: cross-domain plant identification based on herbarium sheets, (ii) BirdCLEF: bird species recognition in audio soundscapes, (iii) GeoLifeCLEF: location-based prediction of species based on environmental and occurrence data and (iv) Snake-CLEF: image-based snake identification.

Diversity

The western Palaearctic species of the hawkmoth genus Hyles (Lepidoptera: Sphingidae) have long been the subject of molecular phylogenetic research. However, much less attention has been paid to the taxa inhabiting the central and eastern Palaearctic, particularly Central Asia, where almost 50% of the species diversity of the genus occurs. Yet, many taxonomic conundrums hinder a proper assessment of the true diversity in these moths. One still unresolved group of species includes Hyles hippophaes and Hyles chamyla. Despite a largely overlapping morphology and ecology, a plethora of infraspecific taxa display some unique divergent characters over a wide geographical area. In this study, we undertook a taxonomic assessment of each population and resolved this species complex using an integrative approach. A combination of new computational techniques (DAISY-II) in comparative morphology and recent advances in DNA extraction methods and sequencing of museum specimens (WISC) alongside m...

Revista Brasileira de Ciência Avícola, 2015

Trends in Ecology & Evolution, 2020

Tropical biomes are the most diverse plant communities on Earth, and quantifying this diversity at large spatial scales is vital for many purposes. As macroecological approaches proliferate, the taxonomic uncertainties in species occurrence data are easily neglected and can lead to spurious findings in downstream analyses. Here, we argue that technological approaches offer potential solutions, but there is no single silver bullet to resolve uncertainty in plant biodiversity quantification. Instead, we propose the use of AI approaches to build a data-driven framework that integrates several data sources-including spectroscopy, DNA sequences, image recognition and morphological data. Such a framework would provide a foundation for improving species identification in macroecological analyses while simultaneously improving the taxonomic process of species delimitation.

Research Ideas and Outcomes, 2022

Natural history collections play a vital role in biodiversity research and conservation by providing a window to the past. The usefulness of the vast amount of historical data depends on their quality, with correct taxonomic identifications being the most critical. The identification of many of the objects of natural history collections, however, is wanting, doubtful or outdated. Providing correct identifications is difficult given the sheer number of objects and the scarcity of expertise. Here we outline the construction of an ecosystem for the collaborative development and exchange of image recognition algorithms designed to support the identification of objects. Such an ecosystem will facilitate sharing taxonomic expertise among institutions by offering image datasets that are correctly identified by their in-house taxonomic experts. Together with openly accessible machine learning algorithms and easy to use workbenches, this will allow other institutes to train image recognition...

Parasitology Research, 2017

Cadavers attract numerous species and genera of Muscidae, both regular elements of carrion insect assemblages, and accidental visitors. Identification of adult Muscidae may be considered difficult, particularly by nonexperts. Since species identification is a vital first step in the analysis of entomological material in any forensic entomology orientated experiment and real cases, various alternative methods of species identification have been proposed. We investigated possibility of semiautomated identification by means of wing measurements as an alternative for classic morphology and DNA-based approaches. We examined genus-level identification success for 790 specimens representing 13 genera of the most common European cadavers visiting Muscidae. We found 99.8% of examined specimens correctly identified to the genus-level. Without error, the following were identified: Azelia, Eudasyphora, Graphomya, Hydrotaea, Musca, Muscina, Mydaea, Neomyia, Polietes, Stomoxys and Thricops. Genus-level misidentifications were found only in Helina and Phaonia. Discrimination of examined material on the species level within Hydrotaea (318 specimens representing eight species) and Muscina (163 specimens representing four species) showed lower, yet still high average identification success, 97.2 and 98.8%, respectively. Our results revealed relatively high success in both genus and species identification of Muscidae of medico-legal importance. Semiautomated identification by means of wing measurements can be used by non-experts and does not require sophisticated equipment. This method will facilitate the identification of forensically relevant muscids in comparison to more difficult and more time-consuming identification approaches based on taxonomic keys or DNA-based methods. However, for unambiguous identification of some taxa, we recommend complementary use of identification keys.

Computational Intelligence and Neuroscience, 2022

Deep neural networks are e cient methods of recognizing image patterns and have been largely implemented in computer vision applications. Object detection has many applications in computer vision, including face and vehicle detection, video surveillance, and plant leaf detection. An automatic ower identi cation system over categories is still challenging due to similarities among classes and intraclass variation, so the deep learning model requires more precisely labeled and high-quality data. In this proposed work, an optimized and generalized deep convolutional neural network using Faster-Recurrent Convolutional Neural Network (Faster-RCNN) and Single Short Detector (SSD) is used for detecting, localizing, and classifying ower objects. We prepared 2000 images for various pretrained models, including ResNet 50, ResNet 101, and Inception V2, as well as Mobile Net V2. In this study, 70% of the images were used for training, 25% for validation, and 5% for testing. e experiment demonstrates that the proposed Faster-RCNN model using the transfer learning approach gives an optimum mAP score of 83.3% with 300 and 91.3% with 100 proposals on ten ower classes. In addition, the proposed model could identify, locate, and classify owers and provide essential details that include ower name, class classi cation, and multilabeling techniques.

Communications in Computer and Information Science, 2020

Conservation efforts to protect biodiversity rely on an accurate identification process. In the case of plant identification, traditional methods used are manual, time-consuming and require a degree of expertise to operate. As a result, there is an increasing interest today for an automated plant identification system. Such a system can help in aiding plant-related education, promoting ecotourism, creating a digital heritage for plant species among many others. We propose a solution using modern convolutional neural network architectures which achieves stateof-the-art performance for plant classification in the wild. An exhaustive set of experiments are performed to classify 112 species of plants from the challenging Indic-Leaf dataset. The best performing model gives Top-1 precision of 90.08 and Top-5 precision of 96.90.

BMC evolutionary biology, 2017

Hundreds of herbarium collections have accumulated a valuable heritage and knowledge of plants over several centuries. Recent initiatives started ambitious preservation plans to digitize this information and make it available to botanists and the general public through web portals. However, thousands of sheets are still unidentified at the species level while numerous sheets should be reviewed and updated following more recent taxonomic knowledge. These annotations and revisions require an unrealistic amount of work for botanists to carry out in a reasonable time. Computer vision and machine learning approaches applied to herbarium sheets are promising but are still not well studied compared to automated species identification from leaf scans or pictures of plants in the field. In this work, we propose to study and evaluate the accuracy with which herbarium images can be potentially exploited for species identification with deep learning technology. In addition, we propose to study ...

Frontiers in Ecology and Evolution

The main objective of this study was to investigate the taxonomic significance of wing phenotypic variations (size and shape) for classifying potter wasps. This is the first study investigating the wing size and shape variations, as well as wing asymmetry, sexual dimorphism, wing integration, and phylogenetic signal analysis of all known Antodynerus species from the Indian subcontinent: A. flavescens, A. limbatus, and A. punctatipennis. We used forewings and hindwings for geometric morphometric analysis, and we proved that each species’ wing had unique size and shape variations, as well as significant right–left wing asymmetry and sexual dimorphism across the Antodynerus species, as verified by discriminant function analysis. Wings of Vespidae are longitudinally folded; based on that, we tested two alternative wing modular hypotheses for evaluating the wing integration, using two subsets organization, such as anterior–posterior (AP) and proximal-distal (PD) wing modular organization...

PLOS ONE, 2018

Works requiring taxonomic knowledge face several challenges, such as arduous identification of many taxa and an insufficient number of taxonomists to identify a great deal of collected organisms. Machine learning tools, particularly convolutional neural networks (CNNs), are then welcome to automatically generate high-performance classifiers from available data. Supported by the image datasets available at the largest online database on ant biology, the AntWeb (www.antweb.org), we propose here an ensemble of CNNs to identify ant genera directly from the head, profile and dorsal perspectives of ant images. Transfer learning is also considered to improve the individual performance of the CNN classifiers. The performance achieved by the classifiers is diverse enough to promote a reduction in the overall classification error when they are combined in an ensemble, achieving an accuracy rate of over 80% on top-1 classification and an accuracy of over 90% on top-3 classification.

Neotropical entomology, 2016

In this study, we assess image analysis techniques as automatic identifiers of three Anastrepha species of quarantine importance, Anastrepha fraterculus (Wiedemann), Anastrepha obliqua (Macquart), and Anastrepha sororcula Zucchi, based on wing and aculeus images. The right wing and aculeus of 100 individuals of each species were mounted on microscope slides, and images were captured with a stereomicroscope and light microscope. For wing image analysis, we used the color descriptor Local Color Histogram; for aculei, we used the contour descriptor Edge Orientation Autocorrelogram. A Support Vector Machine classifier was used in the final stage of wing and aculeus classification. Very accurate species identifications were obtained based on wing and aculeus images, with average accuracies of 94 and 95%, respectively. These results are comparable to previous identification results based on morphometric techniques and to the results achieved by experienced entomologists. Wing and aculeus ...

Scientific Reports, 2020

The accurate identification of biological control agents is necessary for monitoring and preventing contamination in integrated pest management (IPM); however, this is difficult for non-taxonomists to achieve in the field. Many machine learning techniques have been developed for multiple applications (e.g., identification of biological organisms). Some phytoseiids are biological control agents for small pests, such as Neoseiulus barkeri Hughes. To identify a precise biological control agent, a boosting machine learning classification, namely eXtreme Gradient Boosting (XGBoost), was introduced in this study for the semi-automated identification of phytoseiid mites. XGBoost analyses were based on 22 quantitative morphological features among 512 specimens of N. barkeri and related phytoseiid species. These features were extracted manually from photomicrograph of mites and included dorsal and ventrianal shield lengths, setal lengths, and length and width of spermatheca. The results reve...

International journal of information technology, 2019

Plant identification plays a crucial role in sustaining the balance of the environment and protecting the biodiversity of a region. Recognizing different species of plants using conventional methods for conservation purposes is a tedious task. Today there is a cumulative effort made by computer scientists and botanists to automate the entire process of plant identification with leaf being a key feature for distinguishing different species of plants. With the advancement and utilization of relevant technologies like digital cameras, mobile cameras, newer techniques in image processing, pattern recognition, machine learning, automation of this system has been a reality. In this paper we have reviewed the current status of research on computer vision methodologies for taxonomical identification of plants and have also focused on the research challenges such as the diversity of the taxa to be identified, morphological variation in plants belonging to the same species, smaller interspecies variations, the challenges in acquisition of high quality images and standard datasets. The future trends in use of new technologies, creation of standard databases and interdisciplinary aspect of research is also discussed.