i am doing project with diabetic retinopathy image classification and here exudates(yellowish area ) are early signs of diabetic retinopathy ) . what color feature i have to extract to classify images as normal(no diabetic retinopathy) and abnormal

i am doing project with diabetic retinopathy image classification and here exudates(yellowish area ) are early signs of diabetic retinopathy . what color feature i have to extract to classify images as normal(no diabetic retinopathy) and abnormal(diabetic retinopathy). helathy image and exudate image with ground truth image is attached for your reference. i tried with LOCAL delta E function (suggest my image analyst ) but could get as per requirement please help me with sample code.

Answer 1

[Image Analyst]

Try the Color Thresholder app on the Apps tab of the tool ribbon.

[N RAMAKRISHNA]

Please suggest me how color threshold app

[Image Analyst]

1. Go to the tool ribbon and click the Apps tab.
2. Look for the Color Thresholder in the Image Processing apps section and click it.
3. Click the load image button and read in your image file.
4. Select HSV color space.
5. Click and drag the limits in the histograms to start thresholding your image.