Intelligent Image Retrieval from Distorted and Partial Queries for Rapid Mobile Identification of Pests Threatening Food and the Environment

从扭曲和部分查询中进行智能图像检索，以快速移动识别威胁食品和环境的害虫

• 批准号: DP140101075
• 负责人: Prof Yongsheng Gao
• 金额: $ 28.31万
• 依托单位: Griffith University
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Projects
• 财政年份: 2014
• 资助国家: 澳大利亚
• 起止时间: 2014-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DP140101075
• 关键词: Intelligent; Image; Retrieval; Distorted; Partial

Pests and diseases are major threats to the Australian food industry and environmental biosecurity. A rapid and mobile pest information retrieval system is critical to prevent a pest becoming established and devastating the region. However, automated insect image retrieval remains an unsolved challenge in the research community. This project addresses the fundamental problem of distorted and partial image query in cluttered background in order to achieve pest identification at a much earlier on-site stage. The success of this research will not only make a technical breakthrough towards retrieving objects with movable body parts, but also revolutionise the current pest detection and monitoring process.

病虫害是澳大利亚食品工业和环境生物安全的主要威胁。一个快速和移动的害虫信息检索系统是至关重要的，以防止害虫成为建立和破坏该地区。然而，昆虫图像的自动检索仍然是一个未解决的挑战，在研究界。本计画解决在杂乱背景下的失真及部分影像查询的基本问题，以在更早的现场阶段实现害虫辨识。这项研究的成功不仅将在检索具有可移动身体部位的物体方面取得技术突破，而且还将彻底改变目前的害虫检测和监测过程。