基于fpga人体行为检测系统设计-控制工程专业论文.docxVIP

摘 摘 要 Abstract Abstract I I PAGE PAGE IV 摘 要 人体行为的检测与识别在人与计算机之间的信息交换、机器人视觉伺服以及智能 控制等领域都具有十分重要的应用价值。而检测人体行为是此过程中首先需要实现的 目标。如今随着计算机科学和电子科学技术的飞速发展，应用嵌入式设备来实现对人 体行为的实时检测技术不断被研究与开发，成为当下令人关注的热点问题。 基于 FPGA 的应用系统具有体积小，功耗低，速度快等优点，并且用户可以依据 自己的需求来个性化地设计相应的模块，对已完成系统的维护与升级也十分方便。 FPGA 设计技术在图像处理和许多需要高速计算的系统领域中具有广泛的应用背景， 现已成为电子自动化设计中一种重要的设计方法。 本文对人体行为识别与检测方法的发展和研究现状进行了探讨，对几种行为识别 与检测的方法进行了总结与归纳。提出从人体肤色检测算法入手，设计了基于色彩空 间的肤色检测模块，并基于此设计出对人体上半身的头部，左右手臂的识别追踪的方 法。根据对人类面部表情识别方法的研究，对采集到的人体图像进行肤色检测，灰度 归一化，边缘检测等处理，设计实现了对人体面部表情的检测。整个系统的设计在 FPGA 上完成，采用自顶向下的方法进行设计，把人体行为检测系统分成了若干的子 模块，并对每个模块进行了详细的研究与设计，展示了各主要模块的硬件系统架构。 本设计使用 Verilog HDL 语言对系统进行描述，之后进行了仿真和综合。最后在 FPGA 图像开发板上成功完成了调试和验证。 关键词 图像处理 FPGA 行为检测 肤色检测 边缘检测 Abstract The detection and identification of human behavior has a strong and wide application value in the information exchange between human and computer, the visual servo and intelligent control. And the detection of human behavior is the first goal of the main process. Nowadays, with the rapid development of the computer science and electronic technology, it has already become a hot topic of current concern that the research and development of using embedded devices to realize the real-time detection of human behavior. The application system of FPGA has some obvious advantages with small size, low power consumption, high speed, etc. Besides, the user could design the corresponding module according to their own demand, and the following improvements and upgrades of the system design are also quite convenient. Therefore, FPGA technology has a wide application in the field of image processing and many systems required high-speed computing, which has become an essential method in the electronic design automation. This study has discussed the development and research status of the identification and detection methods of human behavior firstly, and some typical methods have been summarized. Through the detection algorithm of human skin, this study has designed of the skin color detection