AttributeError: 'numpy.ndarray' object has no attribute 'append':Image processing example

Question

I will first explain what I want to do. I have an image and I want to store the pixel values for a specific ROI. For that reason I implement the following loop(found in another topic of the site):

pixels = im.load() 

all_pixels = []
for x in range(SpecificWidth):
  for y in range(SpecificHeight):
    cpixel = pixels[x, y]
    all_pixels.append(cpixel)

However, it does not return a SpecificwidthXSpecificHeight matrix but one of length as much as the values. Because I want to keep the size of the matrix of the ROI I implement the following loop(much the same with the previous):

array=np.array(all_pixels)
roi_pixels = np.zeros((SpecificWidth,SpecificHeight))

for i in range(0,array.shape[0],width):
    c_roi_pixels=all_pixels[i]
    roi_pixels.append(c_roi_pixels)

And I have the error as it mentioned in the title.

To use vstack the arrays should have the same size and I am getting the error: ValueError all the input array dimensions except for the concatenation axis must match exactly. — DimKoim
– DimKoim, Commented Jul 29, 2014 at 10:28

Roland Smith · Accepted Answer · 2014-07-29 10:22:13Z

4

In numpy, append is a function, not a method.

So you should use e.g:

roi_pixels = np.append(roi_pixels, c_roi_pixels)

Note that the append function creates and returns a copy! It does not modify the original.

answered Jul 29, 2014 at 10:22

Roland Smith

43.7k3 gold badges69 silver badges98 bronze badges

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

2 Comments

DimKoim Over a year ago

So the problem is due to the fact that I specify a zero matrix?

Roland Smith Over a year ago

No. The problem is that append is not a method of the ndarray class.

ali_m · Accepted Answer · 2014-07-29 14:23:43Z

2

@RolandSmith is absolutely right about the cause of the error message you're seeing. A much more efficient way to achieve what you're trying to do is to convert the whole image to a numpy array, then use slice indexing to get the pixels corresponding to your ROI:

# convert the image to a numpy array
allpix = np.array(im)

# array of zeros to hold the ROI pixels
roipix = np.zeros_like(allpix)

# copy the ROI region using slice indexing
roipix[:SpecificHeight, :SpecificWidth] = allpix[:SpecificHeight, :SpecificWidth]

edited Jul 29, 2014 at 14:23

answered Jul 29, 2014 at 12:45

ali_m

74.6k28 gold badges230 silver badges314 bronze badges

3 Comments

ali_m Over a year ago

The syntax is x[start:stop:step] - you might want to take a look at the documentation here

DimKoim Over a year ago

I erase my comment because I read it before you answer. It works,thanks!

ali_m Over a year ago

As I said, the syntax is [start:stop:step]. If you want the rows from 0 to 79, and the columns from 0 to 104 (in steps of 1), you could use allpix[0:80:1, 0:105:1]. There are two sets of slice indices in the square brackets, since you're indexing into two dimensions. Also remember that Python indexing starts at 0. Since the default start is 0 and the default step size is 1, you can just specify the stop indices for each dimension, i.e. allpix[:80, :105].

Collectives™ on Stack Overflow

AttributeError: 'numpy.ndarray' object has no attribute 'append':Image processing example

2 Answers 2

2 Comments

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related