The below link shows an algorithm I want to try and use and I have a few questions about.
Firstly, I am assuming that when i=0, the first frame of a video is titled caltrain000.
for i = 0:30 imgINumber = i; imgPNumber = i+2;
I also want to know what the below code does? Because when I try and run the code with my own video frames I get the error - "Index exceeds matrix dimensions".
imgI = imgI(:,1:352); imgP = imgP(:,1:352);
Please help with these two questions
No products are associated with this question.
In regard to your second question,
a = a(:,1:3)
redefines a to be all rows of the current matrix, but only columns 1 through 3.
So, say
a = [ 1 2 3 4 5 6; 7 8 9 10 11 12; 13 14 15 16 17 18];
then you get
>> a = a(:,1:3)
a =
1 2 3
7 8 9
13 14 15
When i=0, imgIFile='./caltrain/gray/caltrain000.ras'
The operation is like crop an image. All rows are selected, but only column 1 to 352 (pixel?) are selected.
Do you know why the image need to be cropped like this for motion estimation?
0 Comments