Hands-On Image Processing with Python

Determinant of Hessian (DoH)

The DoH approach is the fastest of all these approaches. It detects the blobs by computing maximums in the matrix of the Determinant of Hessian of the image. The size of blobs does not have any impact on the speed of detection. Both the bright blobs on the dark background and the dark blobs on the bright backgrounds are detected by this approach, but the small blobs are not detected accurately.

The next code block demonstrates how to implement these aforementioned three algorithms using scikit-image:

from numpy import sqrtfrom skimage.feature import blob_dog, blob_log, blob_dohim = imread('../images/butterfly.png') im_gray = rgb2gray(im)log_blobs = blob_log(im_gray, max_sigma=30, num_sigma=10, threshold=.1)log_blobs[:, ...