how to speed up finding which image pixel color is not in a given color list

Purpose: I would like to speed up the process of finding which image pixel value does not contain one of the colors in a given RGB colors table and to map them to another image with _mistakes.png suffix.

Using two for loops to process each pixel individually considering the large size of the image takes long.

import glob

import numpy as np

import os

import cv2

import os.path

# the given list of defined RGB colors.

CLASSES = {

        0: [0, 0, 0],

        1:[255, 0, 0],

        2:[0, 0, 255],

        3:[0, 255, 0], 

        4:[50, 255, 50],

        5:[100, 255, 100]

        }

for image_path in glob.glob("*.png"):

    name = os.path.split(image_path)[-1]

    _name = os.path.splitext(name)[0]

    img = cv2.imread(image_path)

    img_height, img_width, _ = img.shape

    img_mistakes = np.zeros((img.shape))



    color_codes = np.array(list(CLASSES.values()))



    # the following two for loops take so long.

    for row in range(img_height):

        for col in range(img_width):

            if not (img[row,col] == color_codes).all(1).any():

                img_mistakes[row, col] = [200, 200, 200]  # a chosen color



cv2.imwrite(_name + '_mistakes' + '.png', img_mistakes)

asked Jan 1 at 23:35

Majid Azimi

4541623

1

Use vectorized numpy operations, instead of iterating over all the pixels in the interpreter. e.g. you can find all the coordinates of pixels of certain colour using a single call to an optimized function (easily 2 orders of magnitude faster than what you do now), and then use the result to set all those pixels to your chosen colour.

– Dan Mašek
Jan 1 at 23:44

2

Also, consider multi-threading the operation and breaking the image down into regions for each thread to work on.

– PerpetualJ
Jan 1 at 23:47

add a comment |

Using two for loops to process each pixel individually considering the large size of the image takes long.

import glob

import numpy as np

import os

import cv2

import os.path

# the given list of defined RGB colors.

CLASSES = {

        0: [0, 0, 0],

        1:[255, 0, 0],

        2:[0, 0, 255],

        3:[0, 255, 0], 

        4:[50, 255, 50],

        5:[100, 255, 100]

        }

for image_path in glob.glob("*.png"):

    name = os.path.split(image_path)[-1]

    _name = os.path.splitext(name)[0]

    img = cv2.imread(image_path)

    img_height, img_width, _ = img.shape

    img_mistakes = np.zeros((img.shape))



    color_codes = np.array(list(CLASSES.values()))



    # the following two for loops take so long.

    for row in range(img_height):

        for col in range(img_width):

            if not (img[row,col] == color_codes).all(1).any():

                img_mistakes[row, col] = [200, 200, 200]  # a chosen color



cv2.imwrite(_name + '_mistakes' + '.png', img_mistakes)

asked Jan 1 at 23:35

Majid Azimi

4541623

1

Use vectorized numpy operations, instead of iterating over all the pixels in the interpreter. e.g. you can find all the coordinates of pixels of certain colour using a single call to an optimized function (easily 2 orders of magnitude faster than what you do now), and then use the result to set all those pixels to your chosen colour.

– Dan Mašek
Jan 1 at 23:44

2

Also, consider multi-threading the operation and breaking the image down into regions for each thread to work on.

– PerpetualJ
Jan 1 at 23:47

add a comment |

Using two for loops to process each pixel individually considering the large size of the image takes long.

import glob

import numpy as np

import os

import cv2

import os.path

# the given list of defined RGB colors.

CLASSES = {

        0: [0, 0, 0],

        1:[255, 0, 0],

        2:[0, 0, 255],

        3:[0, 255, 0], 

        4:[50, 255, 50],

        5:[100, 255, 100]

        }

for image_path in glob.glob("*.png"):

    name = os.path.split(image_path)[-1]

    _name = os.path.splitext(name)[0]

    img = cv2.imread(image_path)

    img_height, img_width, _ = img.shape

    img_mistakes = np.zeros((img.shape))



    color_codes = np.array(list(CLASSES.values()))



    # the following two for loops take so long.

    for row in range(img_height):

        for col in range(img_width):

            if not (img[row,col] == color_codes).all(1).any():

                img_mistakes[row, col] = [200, 200, 200]  # a chosen color



cv2.imwrite(_name + '_mistakes' + '.png', img_mistakes)

asked Jan 1 at 23:35

Majid Azimi

4541623

Using two for loops to process each pixel individually considering the large size of the image takes long.

import glob

import numpy as np

import os

import cv2

import os.path

# the given list of defined RGB colors.

CLASSES = {

        0: [0, 0, 0],

        1:[255, 0, 0],

        2:[0, 0, 255],

        3:[0, 255, 0], 

        4:[50, 255, 50],

        5:[100, 255, 100]

        }

for image_path in glob.glob("*.png"):

    name = os.path.split(image_path)[-1]

    _name = os.path.splitext(name)[0]

    img = cv2.imread(image_path)

    img_height, img_width, _ = img.shape

    img_mistakes = np.zeros((img.shape))



    color_codes = np.array(list(CLASSES.values()))



    # the following two for loops take so long.

    for row in range(img_height):

        for col in range(img_width):

            if not (img[row,col] == color_codes).all(1).any():

                img_mistakes[row, col] = [200, 200, 200]  # a chosen color



cv2.imwrite(_name + '_mistakes' + '.png', img_mistakes)

python image opencv

asked Jan 1 at 23:35

Majid Azimi

4541623

asked Jan 1 at 23:35

Majid Azimi

4541623

asked Jan 1 at 23:35

Majid Azimi

4541623

asked Jan 1 at 23:35

Majid Azimi

4541623

asked Jan 1 at 23:35

Majid Azimi

4541623

1

Use vectorized numpy operations, instead of iterating over all the pixels in the interpreter. e.g. you can find all the coordinates of pixels of certain colour using a single call to an optimized function (easily 2 orders of magnitude faster than what you do now), and then use the result to set all those pixels to your chosen colour.

– Dan Mašek
Jan 1 at 23:44

2

Also, consider multi-threading the operation and breaking the image down into regions for each thread to work on.

– PerpetualJ
Jan 1 at 23:47

add a comment |

1

Use vectorized numpy operations, instead of iterating over all the pixels in the interpreter. e.g. you can find all the coordinates of pixels of certain colour using a single call to an optimized function (easily 2 orders of magnitude faster than what you do now), and then use the result to set all those pixels to your chosen colour.

– Dan Mašek
Jan 1 at 23:44

2

Also, consider multi-threading the operation and breaking the image down into regions for each thread to work on.

– PerpetualJ
Jan 1 at 23:47

Use vectorized numpy operations, instead of iterating over all the pixels in the interpreter. e.g. you can find all the coordinates of pixels of certain colour using a single call to an optimized function (easily 2 orders of magnitude faster than what you do now), and then use the result to set all those pixels to your chosen colour.

– Dan Mašek
Jan 1 at 23:44

Also, consider multi-threading the operation and breaking the image down into regions for each thread to work on.

– PerpetualJ
Jan 1 at 23:47

add a comment |

1 Answer
1

active

oldest

votes

There is probably an even faster way than this , but it's a start! My money is on @divakar to know it - hint, hint ;-)

#!/usr/local/bin/python3

import numpy as np

import cv2



# Open image into numpy array

im=cv2.imread('start.png')



# Work out how one pixel of each colour we are looking for looks

black = [0,0,0]

blue  = [255,0,0]

red   = [0,0,255]

green = [0,255,0]



# Find all pixels where the 3 RGB values match the sought colour

blacks = np.all(im==black, axis=2)

blues  = np.all(im==blue , axis=2)

reds   = np.all(im==red  , axis=2)

greens = np.all(im==green, axis=2)



# Make empty (black) output array same size as input image

mistakes = np.zeros_like(im)

# Make anything not matching any of our colours into [200,200,200]

mistakes[~(blacks | blues | reds | greens)] = [200,200,200]



# Save result

cv2.imwrite("result.png",mistakes)

start.png

enter image description here

Result:

enter image description here

edited Jan 2 at 13:19

answered Jan 2 at 13:07

Mark Setchell

90.6k779181

Maybe @hpaulj might know a better way to do this?

– Mark Setchell
Jan 7 at 13:58

add a comment |

Your Answer

StackExchange.ifUsing("editor", function () {
StackExchange.using("externalEditor", function () {
StackExchange.using("snippets", function () {
StackExchange.snippets.init();
});
});
}, "code-snippets");

StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "1"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);

StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});

function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});

}
});

draft saved

draft discarded

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53999778%2fhow-to-speed-up-finding-which-image-pixel-color-is-not-in-a-given-color-list%23new-answer', 'question_page');
}
);

Post as a guest

Name

Required, but never shown

1 Answer
1

active

oldest

votes

1 Answer
1

active

oldest

votes

There is probably an even faster way than this , but it's a start! My money is on @divakar to know it - hint, hint ;-)

#!/usr/local/bin/python3

import numpy as np

import cv2



# Open image into numpy array

im=cv2.imread('start.png')



# Work out how one pixel of each colour we are looking for looks

black = [0,0,0]

blue  = [255,0,0]

red   = [0,0,255]

green = [0,255,0]



# Find all pixels where the 3 RGB values match the sought colour

blacks = np.all(im==black, axis=2)

blues  = np.all(im==blue , axis=2)

reds   = np.all(im==red  , axis=2)

greens = np.all(im==green, axis=2)



# Make empty (black) output array same size as input image

mistakes = np.zeros_like(im)

# Make anything not matching any of our colours into [200,200,200]

mistakes[~(blacks | blues | reds | greens)] = [200,200,200]



# Save result

cv2.imwrite("result.png",mistakes)

start.png

enter image description here

Result:

enter image description here

edited Jan 2 at 13:19

answered Jan 2 at 13:07

Mark Setchell

90.6k779181

Maybe @hpaulj might know a better way to do this?

– Mark Setchell
Jan 7 at 13:58

add a comment |

There is probably an even faster way than this , but it's a start! My money is on @divakar to know it - hint, hint ;-)

#!/usr/local/bin/python3

import numpy as np

import cv2



# Open image into numpy array

im=cv2.imread('start.png')



# Work out how one pixel of each colour we are looking for looks

black = [0,0,0]

blue  = [255,0,0]

red   = [0,0,255]

green = [0,255,0]



# Find all pixels where the 3 RGB values match the sought colour

blacks = np.all(im==black, axis=2)

blues  = np.all(im==blue , axis=2)

reds   = np.all(im==red  , axis=2)

greens = np.all(im==green, axis=2)



# Make empty (black) output array same size as input image

mistakes = np.zeros_like(im)

# Make anything not matching any of our colours into [200,200,200]

mistakes[~(blacks | blues | reds | greens)] = [200,200,200]



# Save result

cv2.imwrite("result.png",mistakes)

start.png

enter image description here

Result:

enter image description here

edited Jan 2 at 13:19

answered Jan 2 at 13:07

Mark Setchell

90.6k779181

Maybe @hpaulj might know a better way to do this?

– Mark Setchell
Jan 7 at 13:58

add a comment |

There is probably an even faster way than this , but it's a start! My money is on @divakar to know it - hint, hint ;-)

#!/usr/local/bin/python3

import numpy as np

import cv2



# Open image into numpy array

im=cv2.imread('start.png')



# Work out how one pixel of each colour we are looking for looks

black = [0,0,0]

blue  = [255,0,0]

red   = [0,0,255]

green = [0,255,0]



# Find all pixels where the 3 RGB values match the sought colour

blacks = np.all(im==black, axis=2)

blues  = np.all(im==blue , axis=2)

reds   = np.all(im==red  , axis=2)

greens = np.all(im==green, axis=2)



# Make empty (black) output array same size as input image

mistakes = np.zeros_like(im)

# Make anything not matching any of our colours into [200,200,200]

mistakes[~(blacks | blues | reds | greens)] = [200,200,200]



# Save result

cv2.imwrite("result.png",mistakes)

start.png

enter image description here

Result:

enter image description here

edited Jan 2 at 13:19

answered Jan 2 at 13:07

Mark Setchell

90.6k779181

There is probably an even faster way than this , but it's a start! My money is on @divakar to know it - hint, hint ;-)

#!/usr/local/bin/python3

import numpy as np

import cv2



# Open image into numpy array

im=cv2.imread('start.png')



# Work out how one pixel of each colour we are looking for looks

black = [0,0,0]

blue  = [255,0,0]

red   = [0,0,255]

green = [0,255,0]



# Find all pixels where the 3 RGB values match the sought colour

blacks = np.all(im==black, axis=2)

blues  = np.all(im==blue , axis=2)

reds   = np.all(im==red  , axis=2)

greens = np.all(im==green, axis=2)



# Make empty (black) output array same size as input image

mistakes = np.zeros_like(im)

# Make anything not matching any of our colours into [200,200,200]

mistakes[~(blacks | blues | reds | greens)] = [200,200,200]



# Save result

cv2.imwrite("result.png",mistakes)

start.png

enter image description here

Result:

enter image description here

edited Jan 2 at 13:19

answered Jan 2 at 13:07

Mark Setchell

90.6k779181

edited Jan 2 at 13:19

answered Jan 2 at 13:07

Mark Setchell

90.6k779181

answered Jan 2 at 13:07

Mark Setchell

90.6k779181

answered Jan 2 at 13:07

Mark Setchell

90.6k779181

Maybe @hpaulj might know a better way to do this?

– Mark Setchell
Jan 7 at 13:58

add a comment |

Maybe @hpaulj might know a better way to do this?

– Mark Setchell
Jan 7 at 13:58

Maybe @hpaulj might know a better way to do this?

– Mark Setchell
Jan 7 at 13:58

add a comment |

draft saved

draft discarded

Thanks for contributing an answer to Stack Overflow!

Please be sure to answer the question. Provide details and share your research!

But avoid …

Asking for help, clarification, or responding to other answers.

Making statements based on opinion; back them up with references or personal experience.

To learn more, see our tips on writing great answers.

draft saved

draft discarded

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

Post as a guest

Name

Required, but never shown

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

Name

Required, but never shown

Name

Required, but never shown

This page is only for reference, If you need detailed information, please check here

搜尋此網誌

Bdtjtk