2. Specify the block size, such as 24*32 (divided into 20*20 blocks);

3. The block immediately takes n pixels, assuming that 20 pixels are taken at first, then the average G value of these 20 pixels is defined as G 1, and the average G value G0 of all pixels in the block is calculated;

4. Find the variance varG between G 1 and G0 and store it in the array for later use;

image = im read(' tupian . jpg ')；

g 1 = 0； temp =[]；

Because i= 1:24:480

For j= 1:32:640

area=image(i:i+23，j:j+3 1，:)； % take out the area

For n= 1:20

x = round(rand()* 24)；

y = round(rand()* 32)； % and then generate the x and y coordinates of the point to be taken.

And x==0 | y==0

x = round(rand()* 24)；

y = round(rand()* 32)；

end

g 1 = g 1+double(area(x，y，2))； %G 1 Save the sum of g values of 20 points in this block.

end

g 1 = double(g 1)/20； %G 1 is the average value of g at 20 points.

G0=mean(mean(area(:，:，2))； %G0 saves the average value of the g value of this block.

G=[G 1，G0]；

varG = var(G)； % variance

temp =[temp； varG]；

end

end

The above program can be run.