Extracting retina region from 2D images

Question

I am working on Diabetic Retinopathy detection. The database comprises of 2D images that looks like this one:

Now, I want to extract just the retina part from the images so that the final image looks like this one:

I have tried Canny and HoughCircles but it failed. Is there any way to extract this region using OpenCv or scikit-image or PIL?

Answer 1

Since the foreground is so obvious, you can use simple thresholding to find it and then use regionprops to get the crop you need:

from skimage import io, color, filters, measure

image = io.imread('https://i.sstatic.net/kKBiU.jpg')
grayscale = color.rgb2gray(image)
foreground = (grayscale > filters.threshold_otsu(grayscale)).astype(int)
foreground_properties = measure.regionprops(foreground)[0]
cropped = image[foreground_properties.slice]