| Summary: | 碩士 === 國立海洋大學 === 電機工程學系 === 88 === Abstract
The method of " The Facet Model "simply speaking , is
to put an underlying continuum or piecewise continuous gray level intensity surface on the image . According to the thesis, We can make the conditioning and labeling by way of edge detection , line detection , corner detection and the parameters
of surface as well as the variance of error that are estimated on the facet model .
The first topic of the thesis is about the technology of the facet model on the image analysis . We employ the least-square
error procedure to estimate the parameters of the facet model . Besides , we can tell the peak noise by spatial distribution on
the facet model .
The second topic of the thesis is about the method of edge detection . We first make the edge detection based on gradient , through our statistic G and statistic F . According to the thesis , we use the integrated directional derivative gradient operator to compare with other operators . We can demonstrate them under the input of the steps and ramp edge. Besides , we can get the corner points that we want from the provided image through
the way of corner detection .
|
|---|
