Announcement for Downloading full text filePlease respect the Copyright Act.
All digital full text dissertation and theses from this website are authorized the copyright owners. These copyrighted full-text dissertation and theses can be only used for academic, research and non-commercial purposes. Users of this website can search, read, and print for personal usage. In respect of the Copyright Act of the Republic of China, please do not reproduce, distribute, change, or edit the content of these dissertations and theses without any permission. Please do not create any work based upon a pre-existing work by reproduction, Adaptation, Distribution or other means.
URN etd-0821112-140922 Statistics This thesis had been viewed 2862 times. Download 4281 times. Author Wei-han Huang Author's Email Address No Public. Department Computer Science and Enginerring Year 2011 Semester 2 Degree Master Type of Document Master's Thesis Language zh-TW.Big5 Chinese Page Count 47 Title Integrated Facial Feature with Depth Information for Face Recognition Keyword ASM Face Detection 3D Face Recognition Depth Image Face Recognition Face Recognition Depth Image 3D Face Recognition Face Detection ASM Abstract In recent years, many face recognition systems were proposed, in which color images were utilized. Due to shadow and light reflection, the recognition rates were reduced. In this paper, depth images captured by Microsoft Kinect camera is used to aid the recognition process by overcoming the effects caused by ambient light and reflection. For both the color image and the depth image produced, three types of facial features are extracted. They are statistics of the gradient direction of facial feature points, signature of facial feature points, and nose profile. All these features are integrated with weights set according to experimental results. Then, Nearest Neighbor Classifier is utilized to do the personal identification. There are 40 persons in the database. A color face image and a complete depth image per person is used for training. We tested five times per person and the recognition rate achieves 98% with execution speed 400ms/frame. However, the recognition rate reduced to 63% under variety of illuminations. Advisor Committee Chen-chiung Hsieh - advisor
none - co-chair
none - co-chair
Files Date of Defense 2012-07-11 Date of Submission 2012-08-30