improc ass4: Fixed angles bug in Canny and added subplot titles.

improc/ass4/canny.py

 #!/usr/bin/env python
-from matplotlib.pyplot import imread, imshow, subplot, show
+from matplotlib.pyplot import imread, imshow, subplot, show, axis
 from numpy import arctan2, zeros, append, pi
 from numpy.linalg import norm
 from scipy.ndimage import convolve1d
@@ -29,19 +29,19 @@ def canny(F, s, Tl=None, Th=None):
 
             # Gradient norm and rounded angle
             G[p] = norm(append(Gy[p], Gx[p]))
-            A[p] = int(round(arctan2(Gy[p], Gx[p]) * 4 / pi + 1)) % 4
+            A[p] = (4 - int(round(4 * arctan2(Gy[p], Gx[p]) / pi))) % 4
 
     # Non-maximum suppression
     E = zeros(F.shape)
+    compare = [((-1, 0), (1, 0)), ((-1, 1), (1, -1)), \
+               ((0, 1), (0, -1)), ((1, 1), (-1, -1))]
 
     for y in xrange(F.shape[0]):
         for x in xrange(F.shape[1]):
             g = G[y, x]
-            a = A[y, x]
-            compare = [((y, x - 1), (y, x + 1)), ((y - 1, x - 1), \
-                       (y + 1, x + 1)), ((y - 1, x), (y + 1, x)), \
-                       ((y + 1, x - 1), (y - 1, x + 1))]
-            na, nb = compare[a]
+            a, b = compare[A[y, x]]
+            na = (y + a[1], x + a[0])
+            nb = (y + b[1], x + b[0])
 
             if (not in_image(na, G) or g > G[na]) \
                     and (not in_image(nb, G) or g > G[nb]):
@@ -101,19 +101,24 @@ if __name__ == '__main__':
         # Execute with tracing edges
         E, T = canny(F, s, float(argv[2]), float(argv[3]))
 
-        subplot(131)
+        subplot(131, title='Original image')
         imshow(F, cmap='gray')
-        subplot(132)
+        axis('off')
+        subplot(132, title='Gradient magnitudes')
         imshow(E, cmap='gray')
-        subplot(133)
+        axis('off')
+        subplot(133, title='Thresholds applied')
         imshow(T, cmap='gray')
+        axis('off')
     else:
         # Execute until nn-maximum suppression
         E = canny(F, s)
 
-        subplot(121)
+        subplot(121, title='Original image')
         imshow(F, cmap='gray')
-        subplot(122)
+        axis('off')
+        subplot(122, title='Gradient magnitudes')
         imshow(E, cmap='gray')
+        axis('off')
 
     show()

improc/ass4/gauss.py

@@ -2,7 +2,7 @@
 from numpy import zeros, arange, meshgrid, array, matrix
 from math import ceil, exp, pi, sqrt
 from matplotlib.pyplot import imread, imshow, plot, xlabel, ylabel, show, \
-        subplot, xlim, savefig, axis
+        subplot, xlim, savefig, axis, ticklabel_format
 from mpl_toolkits.mplot3d import Axes3D
 from scipy.ndimage import convolve, convolve1d
 from time import time
@@ -96,11 +96,13 @@ if __name__ == '__main__':
         G = convolve(F, W, mode='nearest')
 
         # Show the original image, kernel and convoluted image respectively
-        subplot(131)
+        subplot(131, title='Original image')
         imshow(F, cmap='gray')
-        plot_kernel(W, subplot(132, projection='3d'))
-        subplot(133)
+        axis('off')
+        plot_kernel(W, subplot(132, projection='3d', title='Kernel'))
+        subplot(133, title='Convoluted image')
         imshow(G, cmap='gray')
+        axis('off')
     elif argv[1] == '1d':
         """Calculate the gaussian kernel using derivatives of the specified
         order in both directions."""
@@ -119,11 +121,13 @@ if __name__ == '__main__':
         W = Fy.T * Fx
 
         # Show the original image, kernel and convoluted image respectively
-        subplot(131)
+        subplot(131, title='Original image')
         imshow(F, cmap='gray')
-        plot_kernel(W, subplot(132, projection='3d'))
-        subplot(133)
+        axis('off')
+        plot_kernel(W, subplot(132, projection='3d', title='Kernel'))
+        subplot(133, title='Convoluted image')
         imshow(G, cmap='gray')
+        axis('off')
     elif argv[1] == 'timer':
         """Time the performance of a 1D/2D convolution and plot the results."""
         if len(argv) < 3:

improc/ass4/report/report.tex

@@ -88,7 +88,7 @@ The result of the \texttt{Gauss} function is shown in figure
 Gaussian kernel and the convolved image.
 
 \begin{figure}[H]
-	\hspace{-5cm}
+	\hspace{-3cm}
 	\includegraphics[scale=.6]{gauss_2d_5.pdf}
 	\caption{The result of \texttt{python gauss.py 2d 5}.}
 	\label{fig:gauss-2d}
@@ -202,12 +202,12 @@ the resulting plot will contain an additional image containing a binary image
 of edges. The thresholds can be specified in the range 0-255, they are scaled
 down by the program to match the image's color range.  An example execution of
 edge detection on the cameraman image using a scale of 2, a lower threshold of
-20 and a higher threshold of 60, can be viewed in figure \ref{fig:canny}.
+50 and a higher threshold of 70, can be viewed in figure \ref{fig:canny}.
 
 \begin{figure}[H]
 	\hspace{-5cm}
-	\includegraphics[scale=.6]{canny_2_20_60.pdf}
-	\caption{The result of \texttt{python canny.py 2 20 60}.}
+	\includegraphics[scale=.6]{canny_2_50_70.pdf}
+	\caption{The result of \texttt{python canny.py 2 50 70}.}
 	\label{fig:canny}
 \end{figure}