Styjun

What is the reason behind water not falling from a bucket at the top of loop?

Deflecting lasers with lightsabers

Why adjustbox needs a tweak of raise=-0.3ex with enumitem?

How to avoid a lengthy conversation with someone from the neighborhood I don't share interests with

Different answers of calculations in LuaLaTeX on local computer, lua compiler and on overleaf

(7 of 11: Fillomino) What is Pyramid Cult's Favorite Shape?

Unlocked Package Dependencies

In-Cabinet (sink base) electrical box - Metal or Plastic?

Any information about the photo with Army Uniforms

Is the EU really banning "toxic propellants" in 2020? How is that going to work?

Is the first page of Novel really that important?

Why interlaced CRT scanning wasn't done back and forth?

Does a bard know when a character uses their Bardic Inspiration?

Skipping same old introductions

What is Modern Vipassana?

What license to choose for my PhD thesis?

What is it exactly about flying a Flyboard across the English channel that made Zapata's thighs burn?

How to transform a function from f[#1] to f[x]

Can you shove a friendly creature?

Can't split a feature as invalid geometry - but "Check Validity" says its valid

Has J.J.Jameson ever found out that Peter Parker is Spider-Man?

Where can I see modifications made to the PATH environment variable by the Go installer

Is an "are" omitted in this sentence

How do people drown while wearing a life jacket?

How to save grayscale image in Pytorch?

How to merge two dictionaries in a single expression?How do I check if a list is empty?How do I check whether a file exists without exceptions?How can I safely create a nested directory?Lazy load of images in ListViewConvert an image to grayscale in HTML/CSSHow do I sort a dictionary by value?How do I auto-resize an image to fit a 'div' container?How do I list all files of a directory?How to vertically align an image inside a div

.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty margin-bottom:0;

0

I want to save grayscale image in Pytorch, each image has four gray values, 0, 60, 120 and 180. I try the following way to save images, but the saved image is not I expected.

for i, (inputs) in enumerate(test_generator):
 pred = modelPl(inputs.float()).detach()
 fig,ax = plt.subplots(1,1,figsize = (5,5))
 ax.imshow(pred[0,:,:], cmap = "gray")
 print(pred.shape)
 torchvision.utils.save_image(pred, saveTestPath + 'img_0.png'.format(i)) 

Output:
torch.Size([400, 400])

Expected image:

However, the saved picture is not correct as follows:

python image computer-vision pytorch

share|improve this question

edited Mar 27 at 1:31

rosefun

asked Mar 27 at 1:26

rosefunrosefun

47522 silver badges1313 bronze badges

add a comment | 

0

I want to save grayscale image in Pytorch, each image has four gray values, 0, 60, 120 and 180. I try the following way to save images, but the saved image is not I expected.

for i, (inputs) in enumerate(test_generator):
 pred = modelPl(inputs.float()).detach()
 fig,ax = plt.subplots(1,1,figsize = (5,5))
 ax.imshow(pred[0,:,:], cmap = "gray")
 print(pred.shape)
 torchvision.utils.save_image(pred, saveTestPath + 'img_0.png'.format(i)) 

Output:
torch.Size([400, 400])

Expected image:

However, the saved picture is not correct as follows:

python image computer-vision pytorch

share|improve this question

edited Mar 27 at 1:31

rosefun

asked Mar 27 at 1:26

rosefunrosefun

47522 silver badges1313 bronze badges

add a comment | 

I want to save grayscale image in Pytorch, each image has four gray values, 0, 60, 120 and 180. I try the following way to save images, but the saved image is not I expected.

for i, (inputs) in enumerate(test_generator):
 pred = modelPl(inputs.float()).detach()
 fig,ax = plt.subplots(1,1,figsize = (5,5))
 ax.imshow(pred[0,:,:], cmap = "gray")
 print(pred.shape)
 torchvision.utils.save_image(pred, saveTestPath + 'img_0.png'.format(i)) 

Output:
torch.Size([400, 400])

Expected image:

However, the saved picture is not correct as follows:

python image computer-vision pytorch

share|improve this question

edited Mar 27 at 1:31

rosefun

asked Mar 27 at 1:26

rosefunrosefun

47522 silver badges1313 bronze badges

for i, (inputs) in enumerate(test_generator):
 pred = modelPl(inputs.float()).detach()
 fig,ax = plt.subplots(1,1,figsize = (5,5))
 ax.imshow(pred[0,:,:], cmap = "gray")
 print(pred.shape)
 torchvision.utils.save_image(pred, saveTestPath + 'img_0.png'.format(i)) 

share|improve this question

edited Mar 27 at 1:31

rosefun

asked Mar 27 at 1:26

rosefunrosefun

47522 silver badges1313 bronze badges

share|improve this question

edited Mar 27 at 1:31

rosefun

asked Mar 27 at 1:26

rosefunrosefun

47522 silver badges1313 bronze badges

asked Mar 27 at 1:26

rosefunrosefun

47522 silver badges1313 bronze badges

asked Mar 27 at 1:26

rosefunrosefun

47522 silver badges1313 bronze badges

1 Answer
1

active

oldest

votes

2

It might be that torchvision.utils.save_image requires values to be in range 0 to 1. Your images have values which are greater than 1 and hence the problem.

You can check this by dividing the tensor by 255 (or some appropriate number). You can also try to set normalize=True and see if it can automatically normalize the data for you.

share|improve this answer

answered Mar 27 at 1:37

Umang GuptaUmang Gupta

4,47211 gold badge2121 silver badges4242 bronze badges

• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  
  Thanks, I solve the problem by dividing the tensor by 255.
  
  – rosefun
  Mar 27 at 1:41
  

  
  
  
  

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');
);

Sign up using Google

Sign up using Facebook

Sign up using Email and Password

Post as a guest

Name

Email

Required, but never shown

StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f55368465%2fhow-to-save-grayscale-image-in-pytorch%23new-answer', 'question_page');

);

Post as a guest

Name

Email

Required, but never shown

2

It might be that torchvision.utils.save_image requires values to be in range 0 to 1. Your images have values which are greater than 1 and hence the problem.

You can check this by dividing the tensor by 255 (or some appropriate number). You can also try to set normalize=True and see if it can automatically normalize the data for you.

share|improve this answer

answered Mar 27 at 1:37

Umang GuptaUmang Gupta

4,47211 gold badge2121 silver badges4242 bronze badges

• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  
  Thanks, I solve the problem by dividing the tensor by 255.
  
  – rosefun
  Mar 27 at 1:41
  

  
  
  
  

add a comment | 

2

It might be that torchvision.utils.save_image requires values to be in range 0 to 1. Your images have values which are greater than 1 and hence the problem.

You can check this by dividing the tensor by 255 (or some appropriate number). You can also try to set normalize=True and see if it can automatically normalize the data for you.

share|improve this answer

answered Mar 27 at 1:37

Umang GuptaUmang Gupta

4,47211 gold badge2121 silver badges4242 bronze badges

• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  
  Thanks, I solve the problem by dividing the tensor by 255.
  
  – rosefun
  Mar 27 at 1:41
  

  
  
  
  

add a comment | 

It might be that torchvision.utils.save_image requires values to be in range 0 to 1. Your images have values which are greater than 1 and hence the problem.

You can check this by dividing the tensor by 255 (or some appropriate number). You can also try to set normalize=True and see if it can automatically normalize the data for you.

share|improve this answer

answered Mar 27 at 1:37

Umang GuptaUmang Gupta

4,47211 gold badge2121 silver badges4242 bronze badges

share|improve this answer

answered Mar 27 at 1:37

Umang GuptaUmang Gupta

4,47211 gold badge2121 silver badges4242 bronze badges

answered Mar 27 at 1:37

Umang GuptaUmang Gupta

4,47211 gold badge2121 silver badges4242 bronze badges

answered Mar 27 at 1:37

Umang GuptaUmang Gupta

4,47211 gold badge2121 silver badges4242 bronze badges