Task 1 of problem 4 complete

problem_4.py

@@ -0,0 +1,68 @@
+import numpy as np
+import cv2, random, sys, math
+
+def xy_to_polar(xy, centre):
+    dx = xy[0] - centre[0]
+    dy = xy[1] - centre[1]
+    r = math.sqrt(dx**2 + dy**2)
+    theta = math.atan2(dy, dx)
+    return [r, theta]
+
+def polar_to_xy(rtheta, centre):
+    dx = rtheta[0] * math.cos(rtheta[1])
+    dy = rtheta[0] * math.sin(rtheta[1])
+    x = dx + centre[0]
+    y = dy + centre[1]
+    return [x, y]
+
+def clip(val, minimum, maximum):
+    return sorted((minimum, val, maximum))[1]
+
+def swirlFilter(img, angle, radius, interpolation="nn"):
+    image_y, image_x, image_channels = img.shape
+    output = np.zeros(img.shape, dtype=np.uint8)
+    img_centre = (image_x//2, image_y//2)
+
+    for y in range(image_y):
+        for x in range(image_x):
+            src_coords = [x,y]
+            dist_from_centre = math.sqrt((x-img_centre[0])**2 + (y-img_centre[1])**2)
+            if dist_from_centre < radius:
+                as_polar = xy_to_polar((x, y), img_centre)
+                as_polar[1] += angle*((1-(dist_from_centre/radius))**1.5)
+                src_coords = polar_to_xy(as_polar, img_centre)
+            for c in range(image_channels):
+                if interpolation == "nn":
+                    input_x = clip(round(src_coords[0]), 0, image_x-1)
+                    input_y = clip(round(src_coords[1]), 0, image_y-1)
+                    output.itemset((y, x, c), img.item(input_y, input_x, c))
+                else:
+                    x1 = clip(math.floor(src_coords[0]), 0, image_x-1)
+                    y1 = clip(math.floor(src_coords[1]), 0, image_y-1)
+                    x2 = clip(math.ceil(src_coords[0]), 0, image_x-1)
+                    y2 = clip(math.ceil(src_coords[1]), 0, image_y-1)
+                    input_x = clip(src_coords[0], 0, image_x-1)
+                    input_y = clip(src_coords[1], 0, image_y-1)
+                    if x2 != x1:
+                        I_y1 = img.item(y1, x1, c)*(x2-input_x)/(x2-x1) + img.item(y1, x2, c)*(input_x-x1)/(x2-x1)
+                        I_y2 = img.item(y2, x1, c)*(x2-input_x)/(x2-x1) + img.item(y2, x2, c)*(input_x-x1)/(x2-x1)
+                    else:
+                        I_y1 = img.item(y1, x1, c)
+                        I_y2 = img.item(y2, x1, c)
+                    if y2 != y1:
+                        I = I_y1*(y2-input_y)/(y2-y1) + I_y2*(input_y-y1)/(y2-y1)
+                    else:
+                        I = I_y1
+                    output.itemset((y, x, c), I)
+
+
+    return output
+
+test_img = cv2.imread("input_2.jpg", cv2.IMREAD_COLOR)
+out_img = swirlFilter(test_img, 3*math.pi/4, 150, "bi")
+
+cv2.imshow("Swirl Filter", out_img) # Show image
+key = cv2.waitKey(0)
+
+if (key == ord('x')):
+    cv2.destroyAllWindows()