An automatic contour tracking procedure is illustrated over sequences of x-ray and echo images. The entire procedure is based on two main steps: (i) a furrow of suitable width is obtained at the image luminance variations by filtering the images with a first order derivative filter and (U) the "true" contour is obtained by dropping an approximate starting contour to the bottom of the furrow. Once the contour is determined on the first image it is used as starting contour to determine the contour on the subsequent image and so on. Therefore, when an approximate starting contour is given on the first image, every contour of the sequence is automatically outlined. As the gradient of Gaussian (GoG) provides a reduced output at two-dimensional discontinuities, a novel first order derivative filter is used to produce the furrow maps. The filter is obtained from the generalization of the first order absolute moment.
Contour tracking over image sequences
PALOMBO, CARLO
1996-01-01
Abstract
An automatic contour tracking procedure is illustrated over sequences of x-ray and echo images. The entire procedure is based on two main steps: (i) a furrow of suitable width is obtained at the image luminance variations by filtering the images with a first order derivative filter and (U) the "true" contour is obtained by dropping an approximate starting contour to the bottom of the furrow. Once the contour is determined on the first image it is used as starting contour to determine the contour on the subsequent image and so on. Therefore, when an approximate starting contour is given on the first image, every contour of the sequence is automatically outlined. As the gradient of Gaussian (GoG) provides a reduced output at two-dimensional discontinuities, a novel first order derivative filter is used to produce the furrow maps. The filter is obtained from the generalization of the first order absolute moment.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.