Biography
Enrollment Date: 2011
Graduation Date：2014
Degree：M.S.
Defense Date：2014.12.24
Advisors：Xiang Xie
Department：Institute of Microelectronics，Tsinghua University
Title of Dissertation/Thesis：Research on Touch Detection Method By Using Stripes and Shadow
Abstract:
The challenge of touch sensitive interaction lies in detecting the touch action，in this paper, an effective method is proposed for detecting the touch operation on flat surface (e.g. tables and walls), with only one camera, by using the stripes of projected surface and the shadow of fingers. Its purpose is to be able to quickly and accurately identify the touch operation in different finger touching directions and angles, which can meet the requirements of real-time interactive interface in people's daily operations, but also conducive to system integration and miniaturization, and laid a foundation for future human-computer interaction systems portable.With the corresponding relationship between pro-cam system and projector plane, a formula can be expressed as a linear relationship between the height of fingers and the offset of interface stripes, when the finger touch the surface as a horizontal direction. So, a sub pixel accuracy stripes detection method is built up to calculate the stripe offset which can identify a finger touching the surface. To determine whether a finger suspending or touching the surface with a tilt angle, we take full use of the feature of fingertip to estimate the finger shadow area caused by hovering fingers. In order to verify the effectiveness of this method, a kind of finger model is set up to calculate the height of the finger by using the shadow with the spatial geometric relationship. To take into account the accuracy and timeliness, a judgment process is proposed as follows: Step1, Homograph acquisition; Step2, looking for the ROI region; Step3, the calculation of stripe offsets; Step4, preliminary touch judgment based on the offset; Step5, the final touch judgment based on the shadow.Additionally, this paper proposes a stable and high precision stripe detection algorithm, and analyzes how to accurately extract the stripes in a variety of noisy environments. For finger shadow processing, the paper set up a spatial geometric model to calculate the height of the finger by using the finger shadow Finally, the method is achieved based on the OPENCV, and a keyboard interface experiments were carried out to verify the accuracy of this method