@MASTERSTHESIS\{IMM1997-01276, author = "J. Sternberg", title = "Scanning and modelling of {3D} objects", year = "1997", keywords = "{3D} modelling, sugar beets, shape properties, laser scanner, triangulation, camera calibration", school = "Informatics and Mathematical Modelling, Technical University of Denmark, {DTU}", address = "Richard Petersens Plads, Building 321, {DK-}2800 Kgs. Lyngby", type = "", url = "http://www2.compute.dtu.dk/pubdb/pubs/1276-full.html", abstract = "The work documented in this thesis is initiated by inspiration from the vision project at Danisco Development Center in Nakskov and the interest for {3D} modelling at {IMM}. Danisco is currently investigating the possibilities for the introduction of machine vision for the purpose of automatic classification, volume estima-tion, and top percentage estimation of sugar beets. Ordinary {2D} image analysis methodologies have been deployed, but for further enhancements the concept of {3D} vision has been introduced. In the initial state of the project it has been of interest to obtain full scale {3D} models of sugar beets for an investigation of geometric shape properties helpful for segmenting larger conglomerates of beets. A primary goal for investigating shape properties of sugar beets is also to estimate the location of the top slice. This report deals with the construction of a laser scanner which is used to obtain {3D} coordinate sets describing an object. The scanner is build using off-the-shelf components and is therefore constructed within a very limited budget. The performance of the apparatus is investigated, and it is deployed in scanning a number of sugar beets. Data handling of the raw data is con-ducted for the formation of range images and to enable the usage of {3D} visualization tools. Finally some preliminary investigations on the shape properties of sugar beets are conducted." }