Done HW5

hw5_release/seam_carving.py

@@ -81,7 +81,14 @@ def compute_cost(image, energy, axis=1):
     paths[0] = 0  # we don't care about the first row of paths
 
     ### YOUR CODE HERE
-    pass
+    # MIT 18.S191 gives two for iterators solution
+    for i in range(1, H):
+        M1 = np.insert(cost[i-1, 0:W-1], 0, 1e10, axis=0)
+        M2 = cost[i-1, :]
+        M3 = np.insert(cost[i-1, 1:W], W-1, 1e10, axis=0)
+        M = np.r_[M1, M2, M3].reshape(3, -1)
+        cost[i] = energy[i] + np.min(M, axis=0)
+        paths[i] = np.argmin(M, axis=0) - 1
     ### END YOUR CODE
 
     if axis == 0:
@@ -119,7 +126,10 @@ def backtrack_seam(paths, end):
     seam[H-1] = end
 
     ### YOUR CODE HERE
-    pass
+    j = end
+    for i in range(H-2, -1, -1):
+        j = paths[i+1, j] + seam[i+1]
+        seam[i] = j
     ### END YOUR CODE
 
     # Check that seam only contains values in [0, W-1]
@@ -149,7 +159,12 @@ def remove_seam(image, seam):
     out = None
     H, W, C = image.shape
     ### YOUR CODE HERE
-    pass
+    # out = np.zeros((H, W-1, C))
+    # for i in range(H):
+    #     idx = seam[i]
+    #     out[i] = np.concatenate((image[i, 0:idx], image[i, idx+1: W]), axis=None).reshape(W-1, -1)
+    out = image[np.arange(W) != seam[:, None]].reshape(H, W-1, C)
+
     ### END YOUR CODE
     out = np.squeeze(out)  # remove last dimension if C == 1
 
@@ -197,7 +212,12 @@ def reduce(image, size, axis=1, efunc=energy_function, cfunc=compute_cost, bfunc
     assert size > 0, "Size must be greater than zero"
 
     ### YOUR CODE HERE
-    pass
+    while out.shape[1] > size:
+        energy = efunc(out)
+        cost, paths = cfunc(out, energy)
+        end = np.argmin(cost[-1]) 
+        seam = backtrack_seam(paths, end)
+        out = remove_seam(out, seam)
     ### END YOUR CODE
 
     assert out.shape[1] == size, "Output doesn't have the right shape"
@@ -224,7 +244,11 @@ def duplicate_seam(image, seam):
     H, W, C = image.shape
     out = np.zeros((H, W + 1, C))
     ### YOUR CODE HERE
-    pass
+    out = np.zeros((H, W+1, C))
+    for i in range(H):
+        idx = seam[i]
+        out[i] = np.concatenate((image[i, 0:idx+1],  image[i, idx: W]), axis=None).reshape(W+1, -1)
+        # out[i] = np.insert(image[i], seam[i], image[i, seam[i]], axis=0)
     ### END YOUR CODE
 
     return out
@@ -265,7 +289,12 @@ def enlarge_naive(image, size, axis=1, efunc=energy_function, cfunc=compute_cost
     assert size > W, "size must be greather than %d" % W
 
     ### YOUR CODE HERE
-    pass
+    while out.shape[1] < size:
+        energy = efunc(out)
+        cost, paths = cfunc(out, energy)
+        end = np.argmin(cost[-1]) 
+        seam = backtrack_seam(paths, end)
+        out = duplicate_seam(out, seam)
     ### END YOUR CODE
 
     if axis == 0:
@@ -392,7 +421,11 @@ def enlarge(image, size, axis=1, efunc=energy_function, cfunc=compute_cost, dfun
     assert size <= 2 * W, "size must be smaller than %d" % (2 * W)
 
     ### YOUR CODE HERE
-    pass
+    seams = find_seams(out, size - W)
+    seams = np.expand_dims(seams, axis=2)
+    for i in range(size - W):
+        out = duplicate_seam(out, np.where(seams == i+1)[1])
+        seams = duplicate_seam(seams, np.where(seams == i+1)[1])
     ### END YOUR CODE
 
     if axis == 0: