• Number of object class: 5 (as shown in Fig)
Feature vector: x=[R G B]T
Step 1: Randomly select 10 points from each class as training data
xkn (k=1, 2,…,5; n=1, 2, .., 10) Step 2: Calculating mean vector μk and
Copyright By PowCoder代写 加微信 powcoder
covariance matrix Sk for each class using its training data xkn.
Step 3: x=[R(i,j) G(i,j) B(i,j)]T for pixel (i, j)
Step 4: Calculate likelihood p(x|k) for all classes (k=1, 2,.., 5). Step 5: Classify the pixel (i,j) to class c, if p(x|c) =max p(x|k) . Step 6: Repeat step 3-5 for all pixels
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com