Splitter error: image dimensions and axes labels must be the same

[ol_nikulina]

Hi there! I am new to KNIME and I’m trying the workflow I found on this forum for image textural features detection. It worked fine with my first set of images but I can’t make it work with the second…

Splitter node gives me an error: “Execute failed: Image dimensions and axes labels must be the same for all images!” However, the images have been resized in Python and are actually all the same. I also tried the “Image Resizer” node, and even after using it, I still get the same problem at “Splitter” node.

My workflow is attached. I couldn’t find the answer in the forum, could someone please give me a hint where the problem could be?

Thanks in advance!

[stelfrich]

Hi @ol_nikulina,

@ol_nikulina

I couldn’t find the answer in the forum, could someone please give me a hint where the problem could be?

I would suggest that we actually take a look at the dimensions of the images: Could you please add an Image Properties node and connect it to the Image Reader (Table) please? Configure the node in the Features tab and select the Dimensions feature and share the output table with us.

I can imagine that the X and Y dimensions of the images are the same but something might be going on in the channel dimensions (thinking about the encoding of pixel types and such…).

Best regards,
Stefan

[ol_nikulina]

Hi @stelfrich,

There’re indeed 5 images that are lacking the third dimension, I will delete them from the dataset. Do you know what could’ve caused it?

Kind regards,
Olesia

[stelfrich]

Hi Olesia,

@ol_nikulina

Do you know what could’ve caused it?

That’s a tough one: Are we talking about color images? Maybe some of them are grayscale images instead? Or what is the third dimension?

Best,
Stefan