基于蚂蚁智能搜索行为的多细胞自动跟踪及其应用研究

项目名称： 基于蚂蚁智能搜索行为的多细胞自动跟踪及其应用研究

项目编号： No.61273312

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： 徐本连

作者单位： 常熟理工学院

项目金额： 80万元

中文摘要： 本项目拟把信息科学中的蚂蚁群智能技术应用到生命科学中的多细胞跟踪领域，通过系统、全面地研究细胞的约束条件，建立较为准确的细胞观测模型；从蚂蚁智能搜索行为出发(基于颜色差异、任务不同、似然强度等启发信息)，分别研究基于联合检测与跟踪的、基于检测与跟踪分立的蚂蚁多细胞跟踪技术，并通过研究具有细胞识别属性的蚂蚁多细胞跟踪器和确定性与随机性相混杂的蚂蚁多细胞跟踪器来自动获取多个细胞动力学特性；为了有效评价多细胞跟踪性能，着重研究科学的、全局的多细胞跟踪性能评价指标，从而最终实现在噪声、杂波、细胞覆盖等诸多不确定因素条件下的多细胞自动跟踪。该项研究成果能为生物学、药理学和病理学的研究提供重要的应用参考价值。

中文关键词： 蚁群系统；细胞跟踪；图像处理；；

英文摘要： This project aims to apply ant swarm intelligence techniques in information science into the field of multi-cell tracking in life science. Through fully and systematically investigating cell constraint conditions, multi-cell observation　equation is modeled appropriately. According to the designed feature of ant intelligent searching behavior (based on such heuristic information as color difference,different tasks assigned, likelihood intensity, etc.）, both jointly and independently detecting/tracking ant multi-cell techniques are studied respectively. Furthermore,we develop several ant multi-cell trackers with cell identification and/or with stochastic and deterministic decisions to achieve the dynamic parameters of multiple cells. To evaluate the effectiveness of multi-cell tracking, a scientific and global performance evaluation index is required to design, and eventually automatic multi-cell tracking with many uncertainties (such as noise, clutter and cell overlap) can be fulfilled. It is believed that our results will contribute to the development of biology, pharmacology and pathology, and provide some crucial application values as well.

英文关键词： ant colony system；cell tracking；image processing；；