`skimage.feature.corner:corner_orientations`

source: /dev/scikit-image/src/skimage/feature/corner.py :1295

Signature

def corner_orientations ( image , corners , mask )

Summary

Compute the orientation of corners.

Extended Summary

The orientation of corners is computed using the first order central moment i.e. the center of mass approach. The corner orientation is the angle of the vector from the corner coordinate to the intensity centroid in the local neighborhood around the corner calculated using first order central moment.

Parameters

image : (M, N) array: Input grayscale image.
corners : (K, 2) array: Corner coordinates as (row, col).
mask : 2D array: Mask defining the local neighborhood of the corner used for the calculation of the central moment.

Returns

orientations : (K, 1) array: Orientations of corners in the range [-pi, pi].

Examples

from skimage.morphology import octagon
from skimage.feature import (corner_fast, corner_peaks,
                             corner_orientations)
square = np.zeros((12, 12))
square[3:9, 3:9] = 1
square.astype(int)
corners = corner_peaks(corner_fast(square, 9), min_distance=1)
corners
orientations = corner_orientations(square, corners, octagon(3, 2))
np.rad2deg(orientations)

✓

Aliases

skimage.feature.corner_orientations