I am reading images with the image writer node and process these images with various other nodes. Neither after the image reader node nor after processing I am able to save the table with images (IMG columns) as an excel file with embedded images. This works fine with png images that I created in another knime workflow. Therefore I tried to convert the IMG column to PNG but this failed as well.
<p> </p>
<p>The problem might be associated with the pixel type. Therefore, I tried to use the image converter node and also changed the pixel type for the output of the image reader (tried all different options), without success.</p>
<p>Maybe this information is useful: During the image reading there was the following warning message in the console:</p>
<p>WARN Image Reader (Table) 0:19 Expected positive value for PhysicalSizeY; got -1.96586E9</p>
<p>Any idea? Thanks in advance. Stefan</p>
</td>
</tr>
</tbody>
can you use the "Renderer to Image" node to try to render the PNG Value to an Image? This is actually a bit of a workaround but might help you for now.
If you could províde a small example workflow demonstrating the problem, I will try to find a more general solution.
I tried to render as suggested - without success. I also converted the tiff image to jpg and load the jpg to the workflow - no success with png conversion as well.
I have attached a little workflow and an example image (jpg conversion since upload of tif is not allowed; however the tif image should be included in the workflow.
I took a look at your workflow and was able to create png images out of the ImgPlus cells and also add them to a xls sheet using the XLS Writer node.
All you need is the ImgPlus to PNG Images node followed by a Split Collection Column node. This will create one column for each image in the collection.
The ImgPlus to PNG Images node creates a collection column out of the input, to allow converting image stacks and time series.
More information about collection cells can be found here: https://www.knime.org/blog/collection-cookbook
I do not really understand the concept of the collection columns at present, but your suggestion worked perfectly. Thanks a lot for the quck response and support.