From b4b34653e454a5470cc21d12570098d771912221 Mon Sep 17 00:00:00 2001
From: Alex <eu7209@wayne.edu>
Date: Thu, 28 Jan 2021 15:42:51 -0500
Subject: [PATCH] Update fuzzycmeans.md

Added an example about fuzzy c means on 2D images.
---
 doc/source/fuzzycmeans.md | 32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)

diff --git a/doc/source/fuzzycmeans.md b/doc/source/fuzzycmeans.md
index f75ef27e..238f3987 100644
--- a/doc/source/fuzzycmeans.md
+++ b/doc/source/fuzzycmeans.md
@@ -45,3 +45,35 @@ M = R.centers
 # memberships is a 20x3 matrix
 memberships = R.weights
 ```
+
+```@example julia Example of fcm on 2D images.
+using TestImages, Clustering, ImageView
+
+# Loads a gray scale image. 
+img = testimage("camera")
+# Convert into a normal Array
+img = Array{Float64,2}(img)
+
+# Reshape that image into a vector.
+Nx, Ny = size(img)
+dat = reshape(img, 1, Nx*Ny)
+
+#=
+Feed in the image, # segments, show iteration info 
+and set max iterations
+=#
+result = fuzzy_cmeans(dat, 5, 2, display=:iter, maxiter=200)
+
+#=
+Take a look at one of the segments and reshape
+the data back into its orginal form. 
+
+This image represents how much each pixel 
+belongs to a given segment. For example here we 
+are looking at segment 1. 
+=#
+img_result = reshape(result.weights[:,1],Nx,Ny)
+
+# View the newly segmented image. 
+imshow(img_result)
+