Bear biometrics: developing an individual recognition technique for sloth bears

IF 1.9 4区生物学 Q1 ZOOLOGY Mammalian Biology Pub Date : 2024-02-15 DOI:10.1007/s42991-023-00396-x

{"title":"Bear biometrics: developing an individual recognition technique for sloth bears","authors":"","doi":"10.1007/s42991-023-00396-x","DOIUrl":null,"url":null,"abstract":"<h3>Abstract</h3> Identifying individual animals, especially in large mammals, is an important goal for wildlife biologists and managers. Bears, occupying diverse habitats, face and experience significant conflict. Among Asian bears, the sloth bear Melursus ursinus (Shaw, 1791; Vulnerable IUCN Red List) is reported vulnerable due to negative interactions with humans, requiring solutions like identifying bear individuals using morphological features. To do so, we used an image-comparison algorithm to evaluate the uniqueness of chest markings using structural similarity index (SSIM) and trained a deep learning model based on the EfficientNet architecture for predicting an individual bear classification. We collected 1567 images (of 144 bears) to examine individual-level differences in chestmark patterns. The comparison yielded 98% accuracy in differentiating chestmarks as a unique pattern for an individual. Subsequently, we trained a circular classification model based on EfficientNet framework using augmented 5628 images for training (80%; of 115 bears), which was validated over 95% for top one and 99% for five individual predictions on 1407 testing images (20%; of 115 bears). The final step involved passing 58 non-augmented images (of 29 out-of-train bears), and the top five predictions of closely similar patterns suggested by the model were then manually compared for similarities in shapes, which suggested whether the image belonged to a new individual. The high accuracy of comparison and classification models suggests the potential applicability of this technique for helping maintain the ex-situ bear database, identifying the conflict individual and estimating bear populations.","PeriodicalId":49888,"journal":{"name":"Mammalian Biology","volume":"57 1","pages":""},"PeriodicalIF":1.9000,"publicationDate":"2024-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Mammalian Biology","FirstCategoryId":"99","ListUrlMain":"https://doi.org/10.1007/s42991-023-00396-x","RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ZOOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Identifying individual animals, especially in large mammals, is an important goal for wildlife biologists and managers. Bears, occupying diverse habitats, face and experience significant conflict. Among Asian bears, the sloth bear Melursus ursinus (Shaw, 1791; Vulnerable IUCN Red List) is reported vulnerable due to negative interactions with humans, requiring solutions like identifying bear individuals using morphological features. To do so, we used an image-comparison algorithm to evaluate the uniqueness of chest markings using structural similarity index (SSIM) and trained a deep learning model based on the EfficientNet architecture for predicting an individual bear classification. We collected 1567 images (of 144 bears) to examine individual-level differences in chestmark patterns. The comparison yielded 98% accuracy in differentiating chestmarks as a unique pattern for an individual. Subsequently, we trained a circular classification model based on EfficientNet framework using augmented 5628 images for training (80%; of 115 bears), which was validated over 95% for top one and 99% for five individual predictions on 1407 testing images (20%; of 115 bears). The final step involved passing 58 non-augmented images (of 29 out-of-train bears), and the top five predictions of closely similar patterns suggested by the model were then manually compared for similarities in shapes, which suggested whether the image belonged to a new individual. The high accuracy of comparison and classification models suggests the potential applicability of this technique for helping maintain the ex-situ bear database, identifying the conflict individual and estimating bear populations.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

熊的生物识别技术：开发懒熊个体识别技术

摘要识别动物个体，尤其是大型哺乳动物个体，是野生动物生物学家和管理人员的一个重要目标。熊的栖息地多种多样，它们面临并经历着严重的冲突。据报道，在亚洲熊中，懒熊Melursus ursinus（Shaw，1791；世界自然保护联盟红色名录易危物种）因与人类的负面互动而易受伤害，这就需要利用形态特征识别熊个体等解决方案。为此，我们使用了一种图像比较算法，利用结构相似性指数（SSIM）来评估胸部标记的独特性，并训练了一个基于 EfficientNet 架构的深度学习模型，用于预测熊的个体分类。我们收集了（144 头熊的）1567 张图像，以检查胸印模式的个体差异。比较结果表明，将胸印区分为个体独特模式的准确率为 98%。随后，我们使用增强的 5628 张图像（占 115 头黑熊的 80%）训练了一个基于 EfficientNet 框架的循环分类模型，并在 1407 张测试图像（占 115 头黑熊的 20%）上验证了该模型对前一个个体预测的准确率超过 95%，对五个个体预测的准确率超过 99%。最后一步是通过 58 张非增强图像（29 只未训练的熊），然后人工比较模型所预测的最接近的前五张图像的形状相似性，以确定该图像是否属于新的个体。对比和分类模型的高准确性表明，这种技术可能适用于帮助维护外来黑熊数据库、识别冲突个体和估计黑熊种群数量。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Mammalian Biology 生物-动物学

CiteScore

3.30

自引率

12.50%

发文量

127

审稿时长

10.1 weeks

期刊介绍： Mammalian Biology (formerly Zeitschrift für Säugetierkunde) is an international scientific journal edited by the Deutsche Gesellschaft für Säugetierkunde (German Society for Mammalian Biology). The journal is devoted to the publication of research on mammals. Its scope covers all aspects of mammalian biology, such as anatomy, morphology, palaeontology, taxonomy, systematics, molecular biology, physiology, neurobiology, ethology, genetics, reproduction, development, evolutionary biology, domestication, ecology, wildlife biology and diseases, conservation biology, and the biology of zoo mammals.