Add files via upload

Author: soumenca

Img3.jpg

@@ -0,0 +1,104 @@
+from scipy.ndimage import filters
+import cv2
+import numpy as np
+import matplotlib.pyplot as plt
+import random
+
+def computeHarrisResponse(im, sigma = 3):
+""" Compute the Harris corner detector response function
+for each pixel in a graylevel image. """
+
+# compute derivatives
+imx = np.zeros(im.shape)
+filters.gaussian_filter(im, (sigma,sigma), (0,1), imx) # Ix = Gx * I
+imy = np.zeros(im.shape)
+filters.gaussian_filter(im, (sigma, sigma), (1,0), imy) # Iy = Gy * I
+
+# compute components of the Harris matrix
+Sxx = filters.gaussian_filter(imx*imx, sigma)
+Sxy = filters.gaussian_filter(imx*imy, sigma)
+Syy = filters.gaussian_filter(imy*imy, sigma)
+
+# compute determinant and trace
+Sdet = Sxx*Syy - Sxy*Sxy
+Str = Sxx + Syy
+
+# compute eigenvalue and eigenvector
+m, n = Sxx.shape
+lamda1 = np.zeros(im.shape)
+lamda2 = np.zeros(im.shape)
+for i in range(m):
+for j in range(n):
+H = np.array([[Sxx[i,j], Sxy[i,j]], [Sxy[i,j], Syy[i,j]]])
+e_vals, e_vecs = np.linalg.eig(H)
+lamda1[i,j] = e_vals[0]
+lamda2[i,j] = e_vals[1]
+
+
+## compute the responce
+R1 = np.minimum(lamda1, lamda2) # The Shi-Tomasi corner response
+R2 = (lamda1*lamda2) - 0.04*(lamda1+lamda2)**2 # Harris corner response
+R3 = Sdet - 0.04*(Str**2) # Another common used Harris corner response
+
+return (R1, R2, R3)
+
+
+def getHarrisPoints(harrisim, min_dist = 5, threshold = 0.03): ## random.uniform(0.02, 0.04)
+""" Return corners from a Harris response image	min_dist is the minimum number 
+of pixels separating corners and image boundary. """
+
+# find top corner candidates above a threshold
+corner_threshold = harrisim.max() * threshold
+harrisim_t = (harrisim > corner_threshold)
+#print harrisim_t.shape
+
+# get coordinates of candidates
+coords = np.array(harrisim_t.nonzero()).T
+#print coords.shape
+
+# ...and their values
+candidate_values = [harrisim[c[0],c[1]] for c in coords]
+
+# sort candidates
+index = np.argsort(candidate_values)
+
+# store allowed point locations in array
+allowed_locations = np.zeros(harrisim.shape)
+allowed_locations[min_dist:-min_dist,min_dist:-min_dist] = 1
+
+# select the best points taking min_distance into account
+filtered_coords = []
+for i in index:
+if allowed_locations[coords[i,0],coords[i,1]] == 1:
+filtered_coords.append(coords[i])
+allowed_locations[(coords[i,0]-min_dist):(coords[i,0]+min_dist),
+(coords[i,1]-min_dist):(coords[i,1]+min_dist)] = 0
+return filtered_coords
+
+
+def plotHarrisPoints(image, filtered_coords):
+""" Plots corners found in image. """
+plt.figure()
+plt.gray()
+plt.imshow(image)
+plt.plot([p[1] for p in filtered_coords],[p[0] for p in filtered_coords],'*')
+plt.axis('off')
+plt.show()
+
+
+def main():
+filename = 'Img3.jpg'
+img = cv2.imread(filename)
+grayimg = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
+im = np.float32(grayimg)
+
+R = (R1, R2, R3) = computeHarrisResponse(im)
+for i in range(len(R)):
+filtered_coords = getHarrisPoints(R[i], 5)
+plotHarrisPoints(im, filtered_coords)
+
+
+
+if __name__== "__main__":
+main()
+