Upload New File

src/methods/Yosra_Jelassi/new_reconstruction.py

+"""A file containing a (pretty useless) reconstruction.
+It serves as example of how the project works.
+This file should NOT be modified.
+"""
+
+
+import numpy as np
+from scipy.signal import convolve2d
+
+from src.forward_model import CFA
+
+
+def new_interpolation(op: CFA, y: np.ndarray) -> np.ndarray:
+    """Performs a simple interpolation of the lost pixels.
+
+    Args:
+        op (CFA): CFA operator.
+        y (np.ndarray): Mosaicked image.
+
+    Returns:
+        np.ndarray: Demosaicked image.
+    """
+    z = op.adjoint(y)
+    if op.cfa == 'bayer':
+        demosaicked_red = np.empty_like(z[:,:,0])
+
+        weights_red = np.array([[0, 1],                      
+                                [0, 0]])   
+
+        for i in range(1, z.shape[0] - 1,2):
+            for j in range(1, z.shape[1] - 1,2):
+                demosaicked_red[i-1:i+1, j-1:j+1] = np.sum(z[i-1:i+1, j-1:j+1,0] * weights_red)
+   
+        demosaicked_green = np.empty_like(z[:,:,1])
+
+        weights_green = np.array([[1, 0],
+                                  [0, 1]]) 
+    
+        for i in range(1, z.shape[0] - 1,2):
+            for j in range(1, z.shape[1] - 1,2):
+                demosaicked_green[i-1:i+1, j-1:j+1] = np.sum(z[i-1:i+1, j-1:j+1,1] * weights_green)/2
+
+        weights_blue = np.array([[0, 0],                      
+                                 [1, 0]])  
+        demosaicked_blue = np.empty_like(z[:,:,2])
+
+        for i in range(1, z.shape[0] - 1,2):
+            for j in range(1, z.shape[1] - 1,2):
+                demosaicked_blue[i-1:i+1, j-1:j+1] = np.sum(z[i-1:i+1, j-1:j+1,2] * weights_blue)
+
+        demosaicked_image = np.dstack((demosaicked_red, demosaicked_green, demosaicked_blue))
+    else:
+        demosaicked_red = np.empty_like(z[:,:,0])
+
+        weights_red = np.array([[0, 0, 1, 1 ],                      
+                                [0, 0, 1, 1 ],
+                                [0, 0, 0, 0 ],
+                                [0, 0, 0, 0 ]])   
+
+        for i in range(2, z.shape[0] - 2, 4):
+            for j in range(2, z.shape[1] - 2, 4):
+                demosaicked_red[i-2:i+2, j-2:j+2] = np.sum(z[i-2:i+2, j-2:j+2,0] * weights_red)/4
+   
+        demosaicked_green = np.empty_like(z[:,:,1])
+
+        weights_green = np.array([[1, 1, 0, 0 ],                      
+                                  [1, 1, 0, 0 ],
+                                  [0, 0, 1, 1 ],
+                                  [0, 0, 1, 1 ]])   
+    
+        for i in range(2, z.shape[0] - 2, 4):
+            for j in range(2, z.shape[1] - 2, 4):
+                demosaicked_green[i-2:i+2, j-2:j+2] = np.sum(z[i-2:i+2, j-2:j+2,1] * weights_green)/8
+
+        weights_blue = np.array([[0, 0, 0, 0 ],                      
+                                 [0, 0, 0, 0 ],
+                                 [1, 1, 0, 0 ],
+                                 [1, 1, 0, 0 ]])  
+        demosaicked_blue = np.empty_like(z[:,:,2])
+
+        for i in range(2, z.shape[0] - 2, 4):
+            for j in range(2, z.shape[1] - 2, 4):
+                demosaicked_blue[i-2:i+2, j-2:j+2] = np.sum(z[i-2:i+2, j-2:j+2,2] * weights_blue)/4
+
+        demosaicked_image = np.dstack((demosaicked_red, demosaicked_green, demosaicked_blue))
+        
+    return demosaicked_image
+
+
+
+def extract_padded(M, size, i, j):
+    N_i, N_j = M.shape
+    res = np.zeros((size, size))
+    middle_size = int((size - 1) / 2)
+
+    for ii in range(- middle_size, middle_size + 1):
+        for jj in range(- middle_size, middle_size + 1):
+            if i + ii >= 0 and i + ii < N_i and j + jj >= 0 and j + jj < N_j:
+                res[middle_size + ii, middle_size + jj] = M[i + ii, j + jj]
+
+    return res
+
+
+def varying_kernel_convolution(M, K_list):
+    N_i, N_j = M.shape
+    res = np.zeros_like(M)
+
+    for i in range(N_i):
+        for j in range(N_j):
+            res[i, j] = np.sum(extract_padded(M, K_list[4 * (i % 4) + j % 4].shape[0], i, j) * K_list[4 * (i % 4) + j % 4])
+
+    np.clip(res, 0, 1, res)
+
+    return res
+
+
+K_identity = np.zeros((5, 5))
+K_identity[2, 2] = 1
+
+K_red_0 = np.zeros((5, 5))
+K_red_0[2, :] = np.array([-3, 13, 0, 0, 2]) / 12
+
+K_red_1 = np.zeros((5, 5))
+K_red_1[2, :] = np.array([2, 0, 0, 13, -3]) / 12
+
+K_red_8 = np.zeros((5, 5))
+K_red_8[:2, :2] = np.array([[-1, -1], [-1, 9]]) / 6
+
+K_red_9 = np.zeros((5, 5))
+K_red_9[:2, 3:] = np.array([[-1, -1], [9, -1]]) / 6
+
+K_red_10 = np.zeros((5, 5))
+K_red_10[:, 2] = np.array([-3, 13, 0, 0, 2]) / 12
+
+K_red_12 = np.zeros((5, 5))
+K_red_12[3:, :2] = np.array([[-1, 9], [-1, -1]]) / 6
+
+K_red_13 = np.zeros((5, 5))
+K_red_13[3:, 3:] = np.array([[9, -1], [-1, -1]]) / 6
+
+K_red_14 = np.zeros((5, 5))
+K_red_14[:, 2] = np.array([2, 0, 0, 13, -3]) / 12
+
+K_list_red = [K_red_0, K_red_1, K_identity, K_identity, K_red_0, K_red_1, K_identity, K_identity, K_red_8, K_red_9, K_red_10, K_red_10, K_red_12, K_red_13, K_red_14, K_red_14]
+
+
+K_green_2 = np.zeros((5, 5))
+K_green_2[2, :] = [-3, 13, 0, 0, 2]
+K_green_2[:, 2] = [-3, 13, 0, 0, 2]
+K_green_2 = K_green_2 / 24
+
+K_green_3 = np.zeros((5, 5))
+K_green_3[2, :] = [2, 0, 0, 13, -3]
+K_green_3[:, 2] = [-3, 13, 0, 0, 2]
+K_green_3 = K_green_3 / 24
+
+K_green_6 = np.zeros((5, 5))
+K_green_6[2, :] = [-3, 13, 0, 0, 2]
+K_green_6[:, 2] = [2, 0, 0, 13, -3]
+K_green_6 = K_green_6 / 24
+
+K_green_7 = np.zeros((5, 5))
+K_green_7[2, :] = [2, 0, 0, 13, -3]
+K_green_7[:, 2] = [2, 0, 0, 13, -3]
+K_green_7 = K_green_7 / 24
+
+K_list_green = [K_identity, K_identity, K_green_2, K_green_3, K_identity, K_identity, K_green_6, K_green_7, K_green_2, K_green_3, K_identity, K_identity, K_green_6, K_green_7, K_identity, K_identity]
+
+
+K_list_blue = [K_red_10, K_red_10, K_red_8, K_red_9, K_red_14, K_red_14, K_red_12, K_red_13, K_identity, K_identity, K_red_0, K_red_1, K_identity, K_identity, K_red_0, K_red_1]
+
+
+
+ker_bayer_red_blue = np.array([[1, 2, 1], [2, 4, 2], [1, 2, 1]]) / 4
+
+ker_bayer_green = np.array([[0, 1, 0], [1, 4, 1], [0, 1, 0]]) / 4
+
+
+