本文为美国密苏里理工大学（作者：MINGZHONG LI）的博士论文，共80页。
研究生物医学体的行为模式有助于科学家了解其潜在机制。利用计算机视觉技术，自动监测可以实现高效和有效的生物医学研究分析。在昆虫群监测、恶性细胞检测与分割、人体器官分割和纳米粒子跟踪等研究领域有着广阔的应用前景。一般来说，计算机视觉技术在生物医学目标监测中的应用包括以下几个阶段：检测、分割和跟踪。每个阶段的挑战都可能导致自动化监测结果不令人满意。这些挑战包括不同的前景背景对比度、快速运动模糊、杂波、目标重叠等。本文研究了每一阶段的挑战，并提出了新的计算机视觉解决方案来克服这些挑战，帮助生物医疗目标在不同环境下实现高精度的自动监控案例。
Studying the behavior patterns ofbiomedical objects helps scientists understand the underlying mechanisms. Withcomputer vision techniques, automated monitoring can be implemented forefficient and effective analysis in biomedical studies. Promising applicationshave been carried out in various research topics, including insect groupmonitoring, malignant cell detection and segmentation, human organ segmentationand nano-particle tracking. In general, applications of computer visiontechniques in monitoring biomedical objects include the following stages:detection, segmentation and tracking. Challenges in each stage will potentiallylead to unsatisfactory results of automated monitoring. These challengesinclude different foreground-background contrast, fast motion blur, clutter,object overlap and etc. In this thesis, we investigate the challenges in eachstage, and we propose novel solutions with computer vision methods to overcomethese challenges and help automatically monitor biomedical objects with highaccuracy in different cases.