项目名称: 生物群体行为语义模型研究及应用
项目编号: No.61272310
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 肖刚
作者单位: 浙江工业大学
项目金额: 79万元
中文摘要: 以水生物群目标行为感知分析为切入点,探索实时性和鲁棒性相融合的生物多目标视觉感知的语义模型,并结合基于水生物行为分析的水质预警系统的理论和应用需求。研究内容包括基于视觉感知的语义模型、多目标行为模型、数字视频分割技术、模式识别以及目标跟踪等学科的相关理论,提出面向水生物的群目标运动检测、跟踪与行为建模分析方法,构建面向城市供水安全的视觉感知生物监测理论体系。以水生物群目标的正常与异常行为差异性分析入手,建立水生物群体目标的运动视觉感知模型,发展多目标指示水生物的运动检测跟踪及行为特征分析方法,构建多目标交互检测与跟踪模型,实现检测跟踪的准确性与实时性。以研究鱼群目标感知行为模式为例,设计出多种水质条件下鱼群目标的视觉感知模型和群体行为语义模型,为实现快速、灵敏、准确的水质预警提供有参考价值的理论。
中文关键词: 鱼群语义行为;水质监测;目标跟踪;;
英文摘要: This project is for the basic application background of real-time detection of water quality and the breakthrough point of behavior perception analysis of water biota, in order to explore a real-time and robust semantic model of visual perception and to meet the application requirement of water quality early warning system.Based on the related theory, such as the semantic perception model, multi-object behavior model, digital video segmentation technology, pattern recognition, object tracking and so on, a method of group target motion detection, tracking and behavior modeling is proposed, for constructing visual perception biological monitoring theory system of the urban water supply security. Through the analysis of the difference between normal and abnormal behavior of water biota, the project will research the visual perception model of water biota group targets, search a analytical method of motion tracking and behavior feature of water biota, construct a model of multi-object interactive detection and tracking, and implement the accuracy and instantaneity of detection and tracking.The emphasis is to research the behavior pattern of fish group, design a visual perception method of normal and abnormal behavior, build semantic model of group behavior and implement a rapid, sensitive and accurate early warning
英文关键词: Semantic behavior of fish;Water quality monitoring;Fish target tracking;;