python - Having used Kirsch's filter for segmentation of blood vessels, I want to remove an unwanted part -
i've used kirsch's filter obtain following result, want remove circular detection (which aren't vessels) i've pointed red arrow below:
how go removing circular detection, not blood vessel? here's code current segmentation:
h1 = np.array([[5, -3, -3], [5, 0, -3], [5, -3, -3]], dtype=np.float32)/ 15 h2 = np.array([[-3, -3, 5], [-3, 0, 5], [-3, -3, 5]], dtype=np.float32) / 15 h3 = np.array([[-3, -3, -3], [5, 0, -3], [5, 5, -3]], dtype=np.float32) / 15 h4 = np.array([[-3, 5, 5], [-3, 0, 5], [-3, -3, -3]], dtype=np.float32) / 15 h5 = np.array([[-3, -3, -3], [-3, 0, -3], [5, 5, 5]], dtype=np.float32) / 15 h6 = np.array([[5, 5, 5], [-3, 0, -3], [-3, -3, -3]], dtype=np.float32) / 15 h7 = np.array([[-3, -3, -3], [-3, 0, 5], [-3, 5, 5]], dtype=np.float32) / 15 h8 = np.array([[5, 5, -3], [5, 0, -3], [-3, -3, -3]], dtype=np.float32) / 15
my desired result (note segmentation of different image):
i try thresholding first image (plain otsu method may suffice) , selecting objects, visible on both images. circle pretty bright, whereas veins (?) dark.
Comments
Post a Comment