Why am I having errors while executing Trainable Segmentation?

nmscanner

I am new to Rapidminer. I am actually using some JPEG magnetic resonance images to perform trainable segmentation but encountered error on the way. Based on the error message, I suspect it is the image I used but not sure at all. I would be very grateful if anyone could point out the right direction to me. Thanks!!

varunm1

I would be very grateful to have more explanation on your kind answer!!

Are you looking related to error? if so, the input you connected to "ROI" port of "Add ROI" operator is not correct. From the image you posted regarding the error, I can see that you used multiply operator to multiply the output of read image and connected the image outputs to the two input ports of Add ROI operator. The add ROI operator is used to specify the region of interest which is the area of an image you are looking to focus on. This  "ROI" port needs an XML file that describes your area of interest in an image rather than an image file (JPED, PNG, etc.) as input. If you dont have any specific region of interest, you donot need to provide input to ""ROI" port of "Add RoI" operator. You can also see the tutorial of "Add ROI" operator to see how the point of interest is generated and stored in XML.



If you are looking for AWS rekognition? The community samples in rapidminer have some examples on how to use AWS for this. Also, you can see below the thread for more details.

https://community.rapidminer.com/discussion/comment/61282#Comment_61282

varunm1

Hello @nmscanner

Looks like the expected input type and the provided input type are not correct. The "Rol" port looks for XML file as input that specifies the region of interest, but you were specifying an image as input to that post, you can remove that connection.

Also, this IMMI (image mining) extension is no longer supported and RM doesn't recommend using this. For images, AWS rekognition is the one that suggested.

nmscanner

Hi varunm1, many thanks for your answer. I would be very grateful to have more explanation on your kind answer!!

nmscanner

@ varunm1, many thanks, my problem is now resolved. I am grateful.