Study on a functional memory-based VLSI system to grow and adapt to the environment

在这项研究中，我们讨论了特定应用的特征提取、特征变换、自适应学习和识别。提出的上述课题的基础技术被应用于多个图像处理应用中。我们还开发了新的算法和体系结构设计，并在软硬件协同设计系统中实现了它们。在智能车辆系统上，我们选择了几种实时图像处理应用，包括车辆检测、速度标志检测和倾斜检测。为了开发自适应学习算法和体系结构，我们研究了基于支持向量机的自适应学习系统，并将该系统应用于具有计算机辅助诊断的医学图像处理系统。

In this study, we discussed feature extraction, feature transformation, adaptive learning and recognition for specific applications. The proposed fundamental technologies for the above subjects were applied to several image processing applications. We also developed new algorithms and architecture designs and implemented them in the software and hardware co-design system. We selected several real time image processing applications which include vehicle detection, speed sign detection, and lean detection on the intelligent vehicle system. So as to develop the adaptive learning algorithm and architecture, we researched the support vector machine based adaptive learning system and applied the proposed system to the medical image processing system with computer aided diagnosis.