立体视觉皮层诱发电位特征提取研究

【作者】 方林；

【导师】 葛霁光；

【作者基本信息】 浙江大学 ， 生物物理学， 2002， 硕士

【摘要】 视觉的研究始终是科学研究的一个重要的方向，而立体视觉是我们的视觉系统的重要的功能组成部分，对立体视觉的研究将对人类认识自我，改造客观世界等产生重要的影响，特别是在机器人的研究，图象处理，仿生学等领域。在过去的研究中，我们实验室已经在立体视觉相关的心理物理和诱发皮层电位方面进行了大量的研究，并取得了一定的研究成果，例如RDS图对的对比度，亮度，复杂的视差区域，颜色等因素对CEP的影响的研究等。同时还发现可能存在视差深度信息相关的CEP特征发放N2波的存在，这在立体视觉相关的CEP的研究中还是第一次，虽然N2波的存在还需要进一步实验的证实，但已经有许多证据表明了它是视差深度相关的CEP发放。 本文对CEP研究过程中的软件和硬件环境进行研究和改进，同时通过改进CEP信号的分析方法，设计了实验者观察刺激信号寸的立体镜支撑系统，包括刺激的呈现，立体镜固定，受试者适应等模块，使得受试者能够舒适地通过立体镜观察CRT上的刺激信号或者宁丁印在纸上的刺激信号。同时还对脑电仪在电路水平上进行了一定的研究，包括信号采集部分、信号放大部分和信号的显示部分等。 在软件方面，主要从三个方面对软件系统进行了改进，包括CEP信号的采集，信号的处理和基于网络的实验室信息的管理。设计的总体思想是软件设计的模块化和实验相关软件的的集成；主要的目标是提高实验的效率，提高软件的功能，易用性和健壮性等。其中实验室资源的管理部分是本文提高实验室管理水平的创新点。通过管理系统，能够对实验室的软件和硬件进行有效的管理和对资源进行有效的分配，从而提高实验室的管理水平。 信号的分析处理以前主要应用平均叠加的方法处理大量CEP实验数据，还使用数字低通椭圆滤波器对信号进行滤波等，数据的平均叠加的次数一般在120左右。本文通过小波变换和独立分量快速分析模式识别的方法对CEP数据进行了分析处理，同时与数字低通椭圆滤波器的处理效果进行了比较，发现用小波对CEP的数据进行滤波处理时，可以将数据的叠加次数从120降低到60次左右，且信号的质量要比以前的处理方法好；但是独立分量快速分析方法效果不很好。更多还原

【Abstract】 Vision research is always a key region of the modern science. When we talk about our vision system, we have to notice that stereoscopic cognition is an important portion of it and any progress of the research will effect many facet of our life, such as self-control robot, image processing and bionics. During the pass time, we have taken a lot of experiments of stereoscopic cognition in psychology and physiology and attain some results, include the influence of RDS(Random Dots Stereogram) contrast, brightness, complexity of disparity area and color to the cognition. And at the same time, we reported that beside the Nl, P2 wave in CEP(Cortical Evoked Potential) , there is a probability of existing of another wave, we call N2, and it sees sensitive to disparity depth of RDS.I build my research work on the pass and improved the hardware and software of experimental system, and explored some new analyzing methods to pick out the CEP signal wrapped in the EEG more efficiently.On hardware, I designed a instrument to let our testee to observe the RDS, shown on Cathode-ray tube(CRT) or printed on paper, through a stereoscope, and never let the stereoscope effect the observation. The instrument includes 3 modules: RDS showing, stereoscope fixing and testee fitting. At the same, we examined the EEG instrument for some problems, included EEG acquisition box(3 pieces circuits), monitor system and amplifying system.On software, I designed a toolbox of experimental application and divided it into 3 part: CEP collecting application, CEP analyzing application and lab source management application. CEP collecting application included 4 main modules, AD card calibration, CEP testing, RDS showing and CEP acquisition. CEP analyzing application’s main function was to extract CEP data from data files for analyzing program. The last one was a new one of our experimental system. It would improve the ability of managing the lab software and hardware resource and take the lab management into a network world.The last work was exploring the advantages of the new signal analyzing methods: wavelet and FastICA on CEP analysis. Our prior method on analyzing CEP was accumulating the CEP and taking the mean of them, and the accumulating times were about 120. In chapter 4, I performed the analyzing work with wavelet and FastICA and find that wavelet was a high efficient way to filter the CEP signal when I set the wavelet and scaling function to Haar, decomposing level to 4, threshold method to fixed to threshold and white-noise structure to scaled white-noise. It took the advantage that reduced the accumulating times to 60 with more smooth signal and less distortion. But FastICA takes no advantage on this facet.更多还原