Akademik Veri Yönetim Sistemi
Proc. 2nd Int. Conf. on Railway Technology, Ajaccio, Fransa, 08 Nisan 2014, ss.1-15, (Tam Metin Bildiri)
Ballast fouling, often associated with deteriorating railroad track performance, refers
to the condition when the ballast layer changes its composition and becomes much
finer in grain size distribution. This paper describes an image analysis approach to
characterize different stages of railroad ballast degradation studied using Los
Angeles abrasion testing in the laboratory. An aggregate image analysis approach is
utilized to investigate ballast particle abrasion and breakage trends at every stage
through detailed quantifications of individual ballast particle size and shape
properties. Aggregate image processing or segmentation techniques have been also
developed and used in this study to analyze the two-dimensional images of ballast
aggregate samples captured by a commonly used DSLR camera in the field for
extraction and analyses of individual aggregate particle size and shape properties.
The segmented individual particle images were fed into the validated University of
Illinois Aggregate Image Analyzer (UIAIA) processing algorithms to compute
particle size and shape properties using the imaging based indices of flat and
elongated ratio (FER), angularity index (AI), and surface texture index (STI). The
performance of the field imaging and segmentation methodology was evaluated by
means of a case study involving field images of railroad aggregate samples collected
from various ballast depths in a mainline freight railroad track. Image analysis
results of ballast particles larger than 9.5 mm (3/8 in.) scanned after a different
number of turns of the LA abrasion drum showed good correlations between percent
changes in aggregate shape properties, i.e., imaging based flatness and elongation,
angularity and surface texture indices, and the fouling index (FI). Such relationships
to be established between in-service track fouling levels and ballast size and shape
properties using similar field imaging techniques would help to better understand
field degradation trends and as a result, improve ballast serviceability and life cycle
performance.