作为计算机视觉、图像处理和模式识别领域最活跃的研究方向之一，运动目标跟踪在军事、自动监控、车辆导航、视频检索、人机交互和机器人等多种领域都有着极其重要的应用价值。本项目拟将生物视觉注意与人脑记忆机制引入到计算机视觉中，探索和研究一种新的运动目标提取和跟踪建模方法。首先基于认知科学研究成果和生物视觉感知机制建立一种视觉信息处理认知模型，用于模拟人类对视觉模式进行感知、注意、匹配、分类、学习和决策的过程；在此基础上基于生物视觉注意机制研究人类视觉对显著特征的感知和注意选择，建立一种动态显著性特征提取模型；然后基于生物视觉稀疏编码机制研究目标特征的表达、存储和提取方法并将其与视觉注意机制相结合用于感兴趣运动目标的提取；最后基于三阶段记忆理论建立一种记忆式目标模板匹配与更新方法并将其用于目标跟踪，以解决复杂场景下目标间的遮挡、目标发生变形以及目标姿态变化等问题。

As one of the most active research directions in the fields of computer vision, image processing, and pattern recognition, moving object tracking has been found vital values in application areas such as military, automated surveillance, vehicle navigation, video indexing, human-computer interaction, and robotics. This project will explore and study a novel method for object extraction and tracking inspired by biological visual attention and brain memorization mechanisms. First, a biologically inspired visual information processing model(BI-VIPM) will be proposed based on cognitive science for mimiking some mechanisms of human vision, such as perception, attention, matching, classification, learning, and decision, etc. Then the conspicuity perception and attention selection mechanism will be studied and a dynamic visual conspicuity feacture extraction model will be built based on biological visual attention mechanism. And then the representation, storing, and retrieving of the objects will be studied based on biological visual sparse coding mechanism and used for extracting the objects of interested combined with the biological visual attention mechanism. Finally, a memory-based object template matching and updating model will be proposed according to the three-stage memory theory and used for object tracking, which will address some hot issues in object tracking, such as object occlusion, object deformation, and object appearance variation, etc.