Interpolation
Interpolation is how new pixel intensities are estimated when an image transformation is applied. It is implemented in SciPy using sets of spline functions.
Editing the interpolation order when using a function such as ndi.zoom() modifies the resulting estimate: higher orders provide more flexible estimates but take longer to compute.
For this exercise, upsample im and investigate the effect of different interpolation orders on the resulting image.
This is a part of the course
“Biomedical Image Analysis in Python”
Exercise instructions
- Use
ndi.zoom()to upsampleimfrom a shape of128, 128to512, 512twice. First, use an interpolationorderof 0, then setorderto 5. - Print the array shapes of
imandup0. - Plot close-ups of the images. Use the index range
128:256along each axis.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Upsample "im" by a factor of 4
up0 = ndi.zoom(____, zoom=____, order=____)
up5 = ____
# Print original and new shape
print('Original shape:', ____)
print('Upsampled shape:', ____)
# Plot close-ups of the new images
fig, axes = plt.subplots(1, 2)
axes[0].imshow(up0[128:256, 128:256])
axes[1].imshow(____)
format_and_render_plots()This exercise is part of the course
Biomedical Image Analysis in Python
Learn the fundamentals of exploring, manipulating, and measuring biomedical image data.
For the final chapter, you'll need to use your brain... and hundreds of others! Drawing data from more than 400 open-access MR images, you'll learn the basics of registration, resampling, and image comparison. Then, you'll use the extracted measurements to evaluate the effect of Alzheimer's Disease on brain structure.
Exercise 1: Spatial transformationsExercise 2: TranslationsExercise 3: RotationsExercise 4: Affine transformExercise 5: Resampling and interpolationExercise 6: ResamplingExercise 7: InterpolationExercise 8: Comparing imagesExercise 9: Mean absolute errorExercise 10: Intersection of the unionExercise 11: Normalizing measurementsExercise 12: Identifying potential confoundsExercise 13: Testing group differencesExercise 14: Normalizing metricsWhat is DataCamp?
Learn the data skills you need online at your own pace—from non-coding essentials to data science and machine learning.