{"id":39082,"date":"2025-07-09T11:35:16","date_gmt":"2025-07-09T03:35:16","guid":{"rendered":"https:\/\/www.1ai.net\/?p=39082"},"modified":"2025-07-09T11:35:16","modified_gmt":"2025-07-09T03:35:16","slug":"%e4%b8%ad%e5%9b%bd%e7%a7%91%e5%ad%a6%e9%99%a2%e6%88%90%e5%8a%9f%e7%a0%94%e5%8f%91%e5%87%ba%e5%ba%95%e6%a0%96%e5%8a%a8%e7%89%a9%e6%99%ba%e8%83%bd%e8%af%86%e5%88%ab%e7%b3%bb%e7%bb%9f%ef%bc%9a%e5%87%86","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/39082.html","title":{"rendered":"The Chinese Academy of Sciences has successfully developed a benthic animal intelligent identification system: the accuracy rate reaches more than 90%, and the whole process is highly automated."},"content":{"rendered":"

July 9, according to the \"Voice of Chinese Academy of Sciences\" public news.Chinese Academy of Sciences<\/a>Utilization of the Institute of Aquatic BiologyAI<\/a>image recognition technology, and successfully developed a benthic animal intelligent recognition system.Realized the whole process of automatic identification and detection of benthic animals<\/strong>.<\/p>\n

\"39f72f55j00sz44lg009sd000v900jdp\"<\/p>\n

1AI learned from the official introduction that benthic animals are an important part of freshwater ecosystems, and their diversity level can reflect the health of the water environment, which is an important indicator for water ecology monitoring.<\/p>\n

However, traditional identification methods based on morphology have many limitations. For example, benthos are morphologically complex and identification requires the examination of subtle features of the target one by one. This makes benthic identification difficult and time-consuming.Difficulty in meeting the highly time-sensitive needs of ecological monitoring<\/strong>The development of rapid and accurate identification techniques and products is therefore urgently needed. Therefore, there is an urgent need to develop rapid and accurate identification techniques and products.<\/p>\n

The team uses two classes of cutting-edge AI image recognition algorithms as the framework for theBuilt a core model for benthic animal identification<\/strong>The YOLO series of target detection algorithms can quickly find and localize detection targets. Among them, the YOLO series of target detection algorithms can quickly find and locate the detection target, while the Vision-Transformer algorithm can accurately distinguish the subtle feature differences between different species by simulating the human attention mechanism.<\/p>\n

On this basis, the team further optimized the model for benthic taxa. With an architecture centered on a multi-scale attention module that allows the model to observe both local details and the overall profile of the target, theImproved capture accuracy of key recognition features<\/strong>. In addition, the use of the overlap recognition algorithm substantially improves the model's detection of benthic animal samples in complex scenarios such as occluded overlapping individuals and high density of small individuals.<\/p>\n

Based on the benthic animal intelligent identification algorithm, the team and the Hubei Institute of Ecological and Environmental Sciences jointly carried out automated mechanical design work and developed the benthic animal intelligent identification system. Relying on hundreds of thousands of high-definition micrographs of benthic animals and strong arithmetic support, the latest model of the system can recognize more than 350 genera and species of benthic animals.Recognition accuracy of more than 90% for common categories<\/strong>.<\/p>\n

The team also equipped the system with professional management software to support project-based benthic sample testing, data interaction and intelligent analysis.<\/p>\n

The operator only needs to put the sample into the fixed container, and the built-in microscopic shooting module of the system will scan the image and recognize the sample intelligently until it generates the data report and inspection report.Highly automated, no human intervention required<\/strong>. The system may bring a new change in benthic ecological monitoring technology.<\/p>","protected":false},"excerpt":{"rendered":"

July 9 news, according to the \"voice of the Chinese Academy of Sciences\" public news, the Chinese Academy of Sciences Institute of Aquatic Biology using artificial intelligence image recognition technology, successfully developed a benthic animal intelligent identification system, realizing the benthic animals throughout the automated identification and detection. 1AI learned from the official introduction that benthic animals are an important part of freshwater ecosystems, and their diversity level can reflect the health of the water environment, which is an important indicator for water ecology monitoring. However, traditional identification methods based on morphology have many limitations. For example, the morphology of benthic animals is complex, and the identification of benthic animals requires the examination of the subtle features of each target one by one. This makes benthic animal identification difficult and time-consuming, and it is difficult to meet the time-sensitive needs of ecological monitoring. Therefore, there is an urgent need to develop fast and accurate identification technologies and products. The team uses two types of cutting-edge artificial intelligence<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[4192,204],"collection":[],"class_list":["post-39082","post","type-post","status-publish","format-standard","hentry","category-news","tag-4192","tag-204"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/39082","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=39082"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/39082\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=39082"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=39082"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=39082"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=39082"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}