Title :
Segmentation of sugar beets using image and graph processing
Author :
Frydendal, Ib ; Jones, Ronald
Author_Institution :
Danisco Sugar Co., Denmark
Abstract :
This paper presents an automatic procedure to segment sugar beets that are lying in a pile. The procedure involves some standard techniques employed by image analysts, including filtering, thresholding, seed selection and water-shedding. We then employ a graph-based approach to find those beets that are lying on top of the pile
Keywords :
computer vision; edge detection; feature extraction; graph theory; image segmentation; mathematical morphology; object recognition; comuter vision; edge detection; feature extraction; filtering; graph nodes; image segmentation; mathematical morphology; object recognition; seed selection; sugar beets; thresholding; water-shedding; Detectors; Filters; Image analysis; Image edge detection; Image processing; Image segmentation; Mathematics; Pixel; Statistics; Sugar industry;
Conference_Titel :
Pattern Recognition, 1998. Proceedings. Fourteenth International Conference on
Conference_Location :
Brisbane, Qld.
Print_ISBN :
0-8186-8512-3
DOI :
10.1109/ICPR.1998.712049
