Method for segmenting object from image data set, involves providing implicit form model for segmenting object from training data records

摘要

The method involves providing an implicit form model for a segmenting object from training data records. The object is segmented by gradual approach of an active contour on base of the form model to the object contour in an image data set. An area of same distance is observed around the active contour superimposing the image data set for repeating the approach. Each screen dot of the image data set is assigned in the area of a signed distance value. The distance value represents a measurement for a distance of the screen dot to the active contour. An independent claim is included for a device for segmenting an object from an image data set for the execution of the method.