Landsat Classification by Team ©

Home
The Problem
The Presentation
The References




The Solution


 

Solution 1. - Histogram matching I.

Step 1. Decompose the image into small cells. Each grid contains 16*16 pixels.
Step 2. Compute the histogram in the RGB levels (All grid has three (red, green, blue) histograms.).
Step 3. Classification based on the correlation of histograms.

Step 4. Segment the image.

 




Solution 2.
- Histogram matching II.

Differences from Histogram Matching I.
Convert the histograms to a greyscale. (Y=0,299 R+0,587 G+0,114 B)








The other steps are the same.


Solution 3.
- Markov Random fields
 

 

 


Solution 4. - Texture-based recognition 

     Features used:

 

                               D   =  no_of_edge_pixels/total_no_edge_pixels

                               E(d) =Sumi,j(|f(i,j)-f(i+d,j)| + |f(i,j)-f(i,j+d)|+|f(i,j)-f(i-d,j)|+|f(i,j)-f(i,j-d)|)

    CD (g1, g2) =  #{((x,y), (x’,y’)): f(x,y)=g1, f(x’,y’)=g2, x=x’+dxi y=y’+dyi}

 


                                  

            Entropy = -  p(i, j)log p(i, j  

  i   j

Variance =  ∑∑ (i - μ)2 p(i, j)   

               i  j

            Total_energy= ∑∑ (p(i, j) ) 

                        i  j

 

 Step 1. Learning

 Step 2. Recognition










Top
Home