Geoscience Reference
In-Depth Information
Since we know the size of a pixel we can now simply count the number of
pixels to estimate the total amount of charcoal in the image. Finally, we
compute the area of all objects, including charcoal.
I7 = im2bw(I4,0.6);
imshow(I7,'XData',[0 ix],'YData',[0 iy]), axis on
xlabel('Centimeters'), ylabel('Centimeters')
title('All Objects')
We are not interested in the absolute areas of charcoal in the image but in the
percentage of charcoal in the sample.
100*sum(sum(I6==0))/sum(sum(I7==0))
ans =
13.4063
h e result suggests that approximately 13% of the sieved sample is charcoal.
As a next step, we could quantify the other components in the sample, such
as ostracods or mineral grains, by choosing dif erent threshold values.
8.12 Shape-Based Object Detection in Images
h e counting of objects within images on the basis of their shapes is a very
time-consuming task. Examples of where this is carried out for round objects
include the counting of planktonic foraminifera shells to infer past sea-
surface temperatures, of diatom frustules to infer past chemical composition
of lake water, and of pollen grains to determine assemblages that can be used
to reconstruct regional air temperature and precipitation. Linear objects that
are determined include faults in aerial photos and satellite images (to derive
the present-day stress i eld of an area) and annual layers (varves) in thin
sections (to establish an annually-resolved sedimentary history).
h e Hough transform, named at er the related 1962 patent of Paul VC
Hough, is a popular technique with which to detect objects within images,
based on their shapes. h e Hough transform was orginally used to detect
linear features, but soon at er being patented it was generalized to identify
objects of any shape (Duda and Hart 1972, Ballard 1981). h e topic by
Gonzalez and others (2009) contains a comprehensive introduction to the
Hough transform and its applications for detecting objects within images.
According to their introduction to the method, the Hough transform is
performed in two steps. In the i rst step an edge detector is used to extract
edge features, such as distinct sediment layers or the outlines of pollen grains,
from an image (Fig. 8.13). In the second step lines (or objects of any other
shape) that trace these edge features are identii ed. h e Image Processing
