Computer Vision Graphics and Image Processing 2008 National






















- Slides: 22

Computer Vision, Graphics, and Image Processing 2008 National Taiwan Ocean University & Fo Guang University Presenter: Jiun-Hao Chen Date: 97/09/04

Outline 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 2

Outline Session D 1 Image & Video Retrieval A content-based retrieval system designed for Chinese ancient bronze images Session D 3 Computer Graphics Radar cross-section visualization Capturing skeleton-based animation data from video Automatic color blending between images 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 3

Session D 1 Image & Video retrieval Nearly noise-free Simple background and uniform illumination Correct position Doesn’t reflect the true size Show the most representative side view Different colors 專業的分類 優:背景單純、亮度適中 缺:相片大小、實際顏色(因受氧化程度不同) 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 4

Session D 1 Image & Video retrieval Polygonal approximation Method Shape descriptor Matching 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 5

Session D 1 Image & Video retrieval Results 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 6

Outline Session D 1 Image & Video Retrieval A content-based retrieval system designed for Chinese ancient bronze images Session D 3 Computer Graphics Radar cross-section visualization Capturing skeleton-based animation data from video Automatic color blending between images 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 7

Session D 3 Computer graphics 傳輸、接收同一部(對隱形機無效) Bi-static or multi-static (雙基地或多基地雷達) 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 8

Session D 3 Computer graphics Method of moment (Mo. M) Shooting and bouncing ray(SBR) 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 9

Session D 3 Computer graphics Data visualization 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 10

Outline Session D 1 Image & Video Retrieval A content-based retrieval system designed for Chinese ancient bronze images Session D 3 Computer Graphics Radar cross-section visualization Capturing skeleton-based animation data from video Automatic color blending between images 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 11

Session D 3 Computer graphics Motivation 人體動作不能套用到動物上面 因此需要利用拍攝動物影片來作動物動作的預估 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 12

Session D 3 Computer graphics ASF file Acclaim Motion Capture System Define skeleton, joints, and bones WGT file How skeleton affects the skin 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 13

Session D 3 Computer graphics Scene estimation Pose estimation Rotation Translation 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 14

Session D 3 Computer graphics Major contribution UI Animation Future work Add example animation data 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 15

Outline Session D 1 Image & Video Retrieval A content-based retrieval system designed for Chinese ancient bronze images Session D 3 Computer Graphics Radar cross-section visualization Capturing skeleton-based animation data from video Automatic color blending between images 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 16

Session D 3 Computer graphics 2008/09/04 source image target image Reinhard et al. this paper Computer Graphics and Interactive Techniques Lab. @NUTN 17

Session D 3 Computer graphics Algorithm RGB to CIE L*A*B* color space Compute means and standard deviations of the three CIE Lab channel Calculate the fuzzy relation values between source and target image Estimate the weight values Transform the colors between source and target image CIE L*A*B* to RGB 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 18

Session D 3 Computer graphics 2008/09/04 source image target image Reinhard et al. this paper Computer Graphics and Interactive Techniques Lab. @NUTN 19

Session D 3 Computer graphics 2008/09/04 source image target image Reinhard et al. this paper Computer Graphics and Interactive Techniques Lab. @NUTN 20

Session D 3 Computer graphics Major contribution Automatically calculate weights according to the fuzzy relations of images Future work Better color measuring method 2008/09/04 Computer Graphics and Interactive Techniques Lab. @NUTN 21

Thank you for attention