Geoscience Reference
In-Depth Information
ans =
0.0951
h e calculated area corresponds approximately to the result from the
grain size analysis. If oversegmentation is a major problem when using
segmentation to count objects in an image, the reader is referred to the topic
by Gonzalez, Woods and Eddins (2009) that describes marker-controlled
watershed segmentation as an alternative method to avoid oversegmentation.
8.11 Quantifying Charcoal in Microscope Images
Quantifying the composition of substances in geosciences, such as the
mineral composition of a rock in thin sections, or the amount of charcoal
in sieved sediment samples, is facilitated by the use of image processing
methods. h resholding provides a simple solution to segmenting objects
within an image that have dif erent coloration or grayscale values. During the
thresholding process, pixels with an intensity value greater than a threshold
value are marked as object pixels (e.g., pixels representing charcoal in an
image) and the rest as background pixels (e.g., all other substances). h e
threshold value is usually dei ned manually through visual inspection of the
image histogram, but numerous automated algorithms are also available.
As an example we analyze an image of a sieved lake-sediment sample
from Lake Nakuru, Kenya (Fig. 8.10). h e image shows abundant light-gray
oval ostracod shells and some mineral grains, as well as gray plant remains
and black charcoal fragments. We use thresholding to separate the dark
charcoal particles and count the pixels of these particles at er segmentation.
At er having determined the number of pixels for all objects distinguished
from the background by thresholding, we use a lower threshhold value to
determine the ratio of the number of pixels representing charcoal to the
number of pixels representing all particles in the sample, i.e., to determine
the percentage of charcoal in the sample.
We read the image of size 1500-by-1500 pixels and assume that the width
and the height of the square image are both one centimeter.
clear
I1 = imread('lakesediment.jpg');
ix = 1; iy = 1;
imshow(I1,'XData',[0 ix],'YData',[0 iy]), axis on
xlabel('Centimeter'), ylabel('Centimeter')
title('Original Image')
h e RGB color image is then converted to a grayscale image using the
function
rgb2gray
.
