Added partial implementation of censure for mode=DoB

skimage/feature/censure.py

@@ -0,0 +1,47 @@
+import numpy as np
+from scipy.ndimage.filters import maximum_filter, minimum_filter
+from skimage.transform import integral_image
+import time
+
+def _get_filtered_image(image, n, mode='DoB'):
+    if mode == 'DoB':
+        inner_wt = (1.0 / (2*n + 1)**2)
+        outer_wt = (1.0 / (12*n**2 + 4*n))
+        integral_img = integral_image(image)
+        filtered_image = np.zeros(image.shape)
+        # TODO : Outsource to Cython
+        start = time.time()
+        for i in range(2 * n, image.shape[0] - 2 * n):
+            for j in range(2 * n, image.shape[1] - 2 * n):
+                inner = integral_img[i + n, j + n] + integral_img[i - n, j - n] - integral_img[i + n, j - n] - integral_img[i - n, j + n]
+                outer = integral_img[i + 2 * n, j + 2 * n] + integral_img[i - 2 * n, j - 2 * n] - integral_img[i + 2 * n, j - 2 * n] - integral_img[i - 2 * n, j + 2 * n]
+                filtered_image[i, j] = outer_wt * outer - (inner_wt + outer_wt) * inner
+        print time.time() - start
+        return filtered_image
+
+
+def censure_keypoints(image, mode='DoB', threshold=0.1):
+    # TODO : Decide mode for convolve function
+    # TODO : Decide number of scales. Image-size dependent?
+    image = np.squeeze(image)
+    if image.ndim != 2:
+        raise ValueError("Only 2-D gray-scale images supported.")
+    # Generating all the scales
+    scale1 = _get_filtered_image(image, 1, mode)
+    scale2 = _get_filtered_image(image, 2, mode)
+    scale3 = _get_filtered_image(image, 3, mode)
+    scale4 = _get_filtered_image(image, 4, mode)
+    scale5 = _get_filtered_image(image, 5, mode)
+    scale6 = _get_filtered_image(image, 6, mode)
+    scale7 = _get_filtered_image(image, 7, mode)
+    # Stacking all the scales in the 3rd dimension
+    scales = np.dstack((scale1, scale2, scale3, scale4, scale5, scale6, scale7))
+    # Suppressing points that are neither minima or maxima in their 3 x 3 x 3
+    # neighbourhood to zero
+    minimas = (minimum_filter(scales, (3, 3, 3)) == scales).astype(int) * scales
+    maximas = (maximum_filter(scales, (3, 3, 3)) == scales).astype(int) * scales
+    # Suppressing minimas and maximas weaker than threshold
+    minimas[np.abs(minimas) < threshold] = 0
+    maximas[np.abs(maximas) < threshold] = 0
+    response = maximas + np.abs(minimas)
+    return response