首页> 外文期刊>Electronics and Communications in Japan / 電子情報通信 >A Communication Aid and Computer Input System Using Only the Human Eye
【24h】

A Communication Aid and Computer Input System Using Only the Human Eye

机译:仅人眼的通信辅助和计算机输入系统

获取原文
获取原文并翻译 | 示例
       

摘要

A communication aid using only the human eye is proposed. A set of candidate characters is displayed on a relatively small computer screen and a light head-mounted display (HMD) that is attached to glasses worn by the user. When the user looks at a candidate character with the left eye, an image of the right eye is taken with a small and light web camera that also is mounted on the glasses. The proposed system can select one character out of 81 characters, with two layers of 9 by 9 character candidate screen images. There is also another selection image including control keys and frequently used sentences. By using image matching between a previously acquired template image for each candidate character and the currently acquired image, the proposed system can determine which of the candidate characters has been selected. By the combined use of blinking and fixed gaze, the proposed system can determine the user-selected key among the candidates. The blink detection method employs a morphologic filter to avoid erroneous pupil detection due to eyebrows and shadows. Thus the user can input sentences. The user can also edit the sentences, which are then read by a text-to-speech (TTS) software tool. Thus the system helps handicapped and disabled persons without voice to communicate with others, because the only function required for conversation is the human eye. In addition, the proposed system can be used as an input system for wearable computing systems. Test results obtained with 6 nondisabled persons showed that the proposed system worked at an acceptable speed of about 1.5 second per character.
机译:提出了仅使用人眼的通信辅助工具。一组候选字符显示在相对较小的计算机屏幕上,并安装在用户佩戴的眼镜上的头戴式显示器(HMD)。当用户用左眼注视候选人物时,用同样安装在眼镜上的小型轻便的网络摄像机拍摄右眼的图像。所提出的系统可以从81个字符中选择一个字符,并具有9乘9字符的两层候选屏幕图像。还有另一个选择图像,包括控制键和常用句子。通过使用针对每个候选字符的先前获取的模板图像与当前获取的图像之间的图像匹配,所提出的系统可以确定已选择了哪个候选字符。通过结合使用眨眼和凝视,所提出的系统可以确定候选者中用户选择的键。眨眼检测方法采用形态过滤器,以避免由于眉毛和阴影而导致错误的瞳孔检测。因此,用户可以输入句子。用户还可以编辑句子,然后由文本到语音(TTS)软件工具读取这些句子。因此,该系统可以帮助没有语音的残障人士和残疾人与他人进行交流,因为对话所需的唯一功能是人眼。另外,所提出的系统可以用作可穿戴计算系统的输入系统。用6名非残障人士获得的测试结果表明,建议的系统以每个字符大约1.5秒的可接受速度工作。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号