Study on a personal digital assistance for the visually impaired using tactile and virtual sound interface

PC界面是基于使用指向设备的gui创建的。然而，对于那些有视觉障碍的人来说，使用它非常困难。视障人士使用电脑的原因主要是为了使用互联网，在网上访问报纸出版公司，阅读报纸的最新文章，以及通过电子邮件与朋友联系。然而，这些任务可以通过使用屏幕阅读器来完成，例如……为了使用因特网，需要一种能高速处理大量信息的新型个人电脑。一般都是像Windows这样的GUI机器，所以视障人士也会使用GUI PC。GUI布局在信息的表达或PC机的操作中起着重要的作用，但按顺序读取字符的屏幕阅读器根本没有考虑到GUI布局。许多视障用户都希望能够识别布局。在这项研究中，我们展示了如何使用一个新的本地化声音系统作为支持系统来识别布局…更多的GUI对象。这些与GUI对象相对应的局部声音是由hrtf在中间平面和两耳之间的耳际差异在水平面上创建的。过滤是由于人体躯干、头部和耳廓的反射和衍射，并由hrtf描述。当听众自己的hrtf被用来合成声音定位线索时，双耳合成效果非常好。利用hrtf和耳机呈现的声音进行双耳合成，进行了计算仿真。我们发现，利用耳廓附近声源的白噪声在垂直方向上采样得到的声音很容易定位。作为一种简单的GUI布局模型，提出了一种25个按钮布局，每个按钮都添加了本地化声音，并检验了本地化声音对抓取按钮位置的有效性。经过深思熟虑，我们发现除了引导鼠标光标外，虚拟声屏还可以用于识别GUI对象的布局。此外，我们还会检查视障人士是否能够自己做出正式的基本图形。我们在本报告中处理的模式仅限于一组行。为了制作正式的图案，我们准备了一个丙烯酸笔输入指南，并将其放在平板电脑上。这个指南有一个9乘9的孔晶格，对应于2维虚拟声音屏幕上9乘9的局部声音。使用这个指南和一支触控笔，视障人士可以制作一个图案。输入模式通过听到局部声音来检查和纠正。实验表明，受试者能够正确输入简单的模式。少

PC interfaces are created based on GUIs which use a pointing device. However it is very hard for those who have visual disabilities to use. The reason that visually impaired persons use PCs is mainly to use the Internet, to access a newspaper publishing company on the Web to read the latest articles of a newspaper, and to contact friends using E-mail. However, these tasks can be done by using a screen reader, e.g.. In order to use the Internet, a new PC which processes a lot of information at high speed is desired. Generally they are GUI machines like Windows, so visually impaired persons will also use a GUI PC. The GUI layout plays an important role in the expression of information or the operation of the PC, but a screen reader which reads characters sequentially does not take into account the GUI layout at all. Many visually impaired users have a desire to recognize the layout.In this study, we show how to use a new localized sound system as a support system to recognize the layout … More of GUI objects. These localized sounds that correspond to the GUI objects are created using by HRTFs in the median plane and interaural differences between both ears in the horizontal plane. The filtering is due to reflections and diffractions from the human torso, head and pinna, and is described by HRTFs. Binaural synthesis works extremely well when the listener's own HRTFs are used to synthesize the sound localization cues. The computational simulation is made by means of binaural synthesis using HRTFs and sound presented by headphones. We found that the sound obtained by sampling in the vertical direction using white noise from a sound source near the pinna was easy to locate. As a simple model of a GUI layout, a 25 button layout with a localized sound added to each button is proposed, and the effectiveness of the localized sounds for grasping the position of the buttons is examined. After much consideration, we found that the virtual sound screen could be applied to recognize the layout of GUI objects in addition to leading a mouse cursor. Also, we examine whether or not a visually impaired person can make a formal basic figure by himself. The patterns that we treat in this report are restricted to a collection of lines. In order to make formal patterns, we prepared an acrylic pen input guide and put it on a tablet. This guide has a 9 by 9 lattice of holes which correspond to 9 by 9 localized sounds on a 2-dimensional virtual sound screen. Using this guide and a stylus pen, the visually impaired can make a pattern. The input pattern is checked and corrected by hearing the localized sound. Experiments show that the subjects can input simple patterns correctly. Less