Abstract: With taking self-portraits users can capture memories, create art, and advance their photography skills. However, most existing camera interfaces are limited in that they do not support device-less remote control, and real-time control over important camera functions, etc. In this paper, a new self-portrait camera system is described and a gesture interface for self-portraits is developed. This self-portrait camera system supports a rotatable front lens and a motion-based gesture interface to control camera functions including the shutter trigger, color balance, and exposure. The gesture-recognition accuracy is experimentally evaluated and the effectiveness of the system is examined in a user study. The results suggest that the proposed gesture recognition technique achieves 85% accuracy, and the gesture-based interface is effective for controlling self-portrait camera options and improves the user experience when taking self-portraits.

Key words: self-portrait, gesture interface, motion gesture, human computer interaction, user interface

摘要： 随着智能移动终端的发展及摄像镜头的小型化，自拍变得越来越流行。如何设计新型自拍交互方法使得用户在自拍过程中能够自由、实时地控制相机是自拍相机交互界面的关键问题。提出利用基于视觉的运动手势交互界面的新方法，使自拍过程中用户只要挥一挥手臂就可以实现与自拍相机的交互功能。使用手势交互的方法，用户可以把相机放在任意的平台上，自由地摆出各种自拍姿态，增加了自拍的丰富性，提高了用户体验。主要提出挥手及画圈两种交互手势，通过组合应用可以实现丰富高效的自拍交互控制功能，如快门控制、白平衡，曝光度等。手势的识别利用相机摄像的实时图像进行处理，采用稀疏光流算法来识别运动手势。用户评估实验表明，所提出运动手势自拍交互界面具有较好的交互效率以及良好的用户满意度，两种手势的识别效率约为85%。

关键词: 自拍, 手势界面, 运动手势, 人机交互, 用户界面