Hallo
I am extracting features from 2D x-rays. The segmentation model I used created png masks of 256x256 pixels. My original x-rays are 1024x1024 pixels. So I have increased the masks with a simple python function, np.repeat(), in both dimensions and converted the array back to png images (with uint8 data type). But when I try to used these masks with the corresponding x-ray (image and mask formats and sizes are the same), I get the following error.
RuntimeError: Exception thrown in SimpleITK LabelStatisticsImageFilter_Execute: d:\a\1\sitk\code\common\include\sitkDualMemberFunctionFactory.hxx:214:
sitk::ERROR: Pixel type: vector of 8-bit unsigned integer is not supported in 2D byclass itk::simple::LabelStatisticsImageFilter
How can I increase a 256x256 mask to 1024x1024 without getting this error?
Thanks for your help!