this is for holding javascript data
Daniel Stanley Tan edited subsection_Identifying_the_Infected_Clusters__.tex
over 8 years ago
Commit id: 3089f4ac55b7fecaa26d1eef4d349680e3a2b57b
deletions | additions
diff --git a/subsection_Identifying_the_Infected_Clusters__.tex b/subsection_Identifying_the_Infected_Clusters__.tex
index 0e694ed..5923a5e 100644
--- a/subsection_Identifying_the_Infected_Clusters__.tex
+++ b/subsection_Identifying_the_Infected_Clusters__.tex
...
\subsection{Identifying the Infected Clusters}
The
function of the K-Means algorithm is
able to cluster similar pixels
together, thus separating together. By doing so, it is able to separate the infected pixels from the healthy pixels. However, the algorithm is oblivious as to which clusters contain infected pixels and which clusters contain healthy pixels.
This information is necessary for the computation of the infestation level. The
task problem now
becomes changes from a segmentation problem to a classification problem.
A classifier requires a set of features that represent the data points and discriminate between their classes. In this case, the data points are the clusters of pixels obtained from the K-Means algorithm. We observed that humans rely on color to distinguish the infected part of the fruit. Therefore, it is logical to choose color as the main feature for the classifier. We transformed the clusters into feature vectors to be used in training the classifier. The feature vector is composed of the centroids $a$ and $b$ (in L*a*b* color space computed from the K-Means algorithm), and the average values of red, green, and blue ($\mu_r$, $\mu_g$, $\mu_b$ in the RGB color space) taken over all the pixels within cluster.