基于眼动跟踪的智能家居控制器

发布时间：2018-05-09 11:09

本文选题：眼动跟踪 + 角膜瞳孔反射法　；参考：《哈尔滨理工大学》2017年硕士论文

【摘要】：眼动跟踪是一种新的人机交互方式,其原理是通过检测人眼的状态与特征来定位视线,并通过眼部动作实现控制。随着智能家居的快速发展与普及,将眼动跟踪技术融合进智能家居也成为近些年的研究热点。基于眼动跟踪的交互方式具有非接触、低干扰等优点,在智能家居控制领域,将会有广阔的应用前景。本文设计的基于眼动跟踪的智能家居控制器,依据角膜瞳孔反射原理,采用SOPC技术,选取FPGA作为控制核心,并结合图像处理技术,实现对智能家居的眼动控制。系统算法设计方面,首先设计了眼部识别算法,在人脸红外图像中识别并定位眼部区域;其次设计了视线识别算法,其对眼部灰度图像依次进行图像分割、边缘检测、椭圆拟合等处理,分别求出瞳孔和光斑的中心坐标;之后设计了视点映射算法,建立人脸图像坐标系与显示屏坐标系间的映射函数,计算出视点在显示屏上的显示位置;此外本文还设计了眨眼检测算法,用来实现人机交互。功能模块设计方面,设计了图像采集模块,图像显示模块,通信模块和SOPC模块。图像采集模块包含红外照明单元和摄像头模组。红外照明单元提供近红外光,摄像头模组为系统提供视频流数据。图像显示模块使用集成了D/A的VGA显示器来显示视点信息和交互界面;以SDRAM作为显存,用来存储图像信息。通信模块通过GPIO与FPGA相连,以红外线为载体发送控制信号。SOPC模块作为系统的核心,以Quartus II开发环境提供的Nios II内核作为CPU,实现图像处理算法。系统测试结果表明,本文设计的基于眼动跟踪的智能家居控制器系统资源占用率低至11%,准确度超过90%,不仅实现眼动跟踪与控制,还留有足够的系统资源来进行后续的功能开发,达到了精度高、可扩展的设计要求。
[Abstract]:Eye movement tracking is a new human-computer interaction method. The principle of eye tracking is to locate the line of sight by detecting the state and characteristics of human eyes and to realize the control by eye movement. With the rapid development and popularity of smart home, eye tracking technology has become a research hotspot in recent years. The interaction mode based on eye movement tracking has the advantages of non-contact and low interference, so it will have a broad application prospect in the field of intelligent home control. The intelligent home controller based on eye movement tracking is designed in this paper. According to the principle of corneal pupillary reflection, SOPC technology is adopted, FPGA is selected as the control core, and image processing technology is combined to realize eye movement control of smart home. In the aspect of system algorithm design, first, the eye recognition algorithm is designed to identify and locate the eye region in the face infrared image; secondly, the line of sight recognition algorithm is designed, which carries on the image segmentation and edge detection to the grayscale image of the eye in turn. The center coordinates of pupil and spot are obtained by ellipse fitting, then the view mapping algorithm is designed, and the mapping function between face image coordinate system and display screen coordinate system is established, and the display position of view point on the display screen is calculated. In addition, a blinking detection algorithm is designed to realize human-computer interaction. Function module design, image acquisition module, image display module, communication module and SOPC module. The image acquisition module includes infrared illumination unit and camera module. Infrared illumination unit provides near-infrared light, camera module provides video stream data for the system. The image display module uses VGA display integrated with D / A to display view information and interactive interface, and uses SDRAM as display memory to store image information. The communication module is connected with FPGA through GPIO. The control signal is sent by infrared ray. SOPC module is used as the core of the system, and the Nios II kernel provided by Quartus II development environment is used as the core of the system to realize the image processing algorithm. The system test results show that the system resource occupancy rate of the intelligent home controller based on eye movement tracking is as low as 11 percent and the accuracy is more than 90. The system not only realizes eye movement tracking and control, but also leaves enough system resources for subsequent functional development. Meet the design requirements of high precision and expandable.
【学位授予单位】：哈尔滨理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TU855

【相似文献】