Civil Engineering Reference
In-Depth Information
Figure 10.3 Concrete column detection using shape and texture (Zhu et al., 2010)
visual signature is then calculated and compared with the signatures of concrete
image samples, which are pre-defined in a materials database (Brilakis and
Soibelman, 2005). If the signature matches concrete, that region is identified as
a concrete column. Column surfaces sharing a long vertical line are assumed to
belong to one column.
Scale/affine-invariant feature-based methods detect an object using image scale/
affine-invariant features. The process of these methods is divided into two stages.
The first phase is to extract scale/affine-invariant features in an image using a
feature detector such as SIFT (Cornelis and Van-Gool, 2008) or RIFT (Lazebnik
Search WWH ::
Custom Search