Premium
3D Image Segmentation of Aggregates from Laser Profiling
Author(s) -
Kim Hyoungkwan,
Haas Carl T.,
Rauch Alan F.,
Browne Craig
Publication year - 2003
Publication title -
computer‐aided civil and infrastructure engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.773
H-Index - 82
eISSN - 1467-8667
pISSN - 1093-9687
DOI - 10.1111/1467-8667.00315
Subject(s) - pixel , canny edge detector , artificial intelligence , computer vision , computer science , binary image , segmentation , maxima and minima , digital image , watershed , binary number , image segmentation , edge detection , pattern recognition (psychology) , image processing , mathematics , image (mathematics) , mathematical analysis , arithmetic
In digital imaging of discrete particles, such as stone aggregates, segmentation is one of the most important processes. Automated scanners of different designs use cameras or lasers to obtain digital images of groups of aggregate particles. To accurately determine particle size and shape parameters, each particle region in the image must be isolated and processed individually. Here, a method for segmenting a particle image acquired from laser profiling is developed using a Canny edge detector and a watershed transformation. Canny edges with rigorous and liberal threshold values are used to outline particle boundaries on a binary image and to check the validity of watersheds, respectively. To find appropriate regional minima in the watershed transformation, a varying search window method is used, where the number of neighboring pixels being compared with the pixel of interest is determined from the height value of the pixel. Test results with this method are promising. When implemented in automated systems that are designed to rapidly assess size and shape characteristics of stone particles, this technique can not only reduce the amount of time required for aggregate preparation, but also increase the accuracy of analysis results.