filtering 2D images in numpy array and calculate in python

Question

I imported an image with three bands. And I entered each band into a numpy array.

Now I try to modify the value of band 1, conditional on band 3.

However, my image has many zero values and must be computed with the exception of zero to speed up the operation.

I think it is faster to find values after excluding the 0 value.

Below is the code I used to do.

cols = 0 
rows = 0
[cols,rows] = test.shape
i= 0
i2 = 0

while i < cols:
    k = 0
    k2 =0
    while k <rows:
        if 0.15>test[i,k]>0.05089 and  30> test3[i,k]>29.8  :
            test[i,k] = 1
....

You may want to consider giving an example. Right now your use of pronouns makes it very difficult to understand what you want. — Mad Physicist
– Mad Physicist, Commented May 6, 2018 at 14:23
For example, I want to find a value of 1 in the array and double it. — 송준석
– 송준석, Commented May 6, 2018 at 14:24
Corrected the contents. If you do not understand, please comment again. — 송준석
– 송준석, Commented May 6, 2018 at 14:32
Your example is bad and you clearly haven't tried running it yourself. You can't start a variable name with a number, and you can just do img *= 2. Please show something that can actually be run. Consider an example involving addition. — Mad Physicist
– Mad Physicist, Commented May 6, 2018 at 14:35

Peter · Accepted Answer · 2018-05-06 20:08:18Z

2

Well it looks like what you want is to select a "mask" and assign to it. Your example is a bit strange and incomplete, but you could achieve what I think you're intending to achieve by replacing the loop with:

test[(0.15>test) & (test>0.05089) & (30>test3) & (test3>29.8)] = 1

What's going on here:

(0.15>test) create a boolean array the same size as test with all elements < 0.15 set to True and the rest False.
The & operators do an elementwise logical AND with the other boolean arrays to make a new boolean array (also the same size as test and test3)
test[XXX] = 1 means "take all elements of XXX which are true, and set the corresponding elements of test to 1" (it's assumed that XXX has the same shape as test (or can be broadcasted to the same shape))

answered May 6, 2018 at 20:08

Peter

13.8k11 gold badges82 silver badges99 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

송준석 Over a year ago

Thank you for your interest in this question.

Collectives™ on Stack Overflow

filtering 2D images in numpy array and calculate in python

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related