added laplacian_filter file

digital_image_processing/filters/laplacian_filter.py

@@ -0,0 +1,96 @@
+# @Author  : ojas-wani
+# @File    : laplacian_filter.py
+# @Time    : 10/04/2023
+
+from cv2 import (
+    imread,
+    cvtColor,
+    COLOR_BGR2GRAY,
+    filter2D,
+    imread,
+    imshow,
+    waitKey,
+    CV_64F,
+    BORDER_DEFAULT,
+)
+import numpy as np
+from gaussian_filter import gaussian_filter
+
+
+def my_laplacian(
+    src, ddepth=-1, ksize=3, scale=1, delta=0, borderType="default"
+) -> np.ndarray:
+    """
+    :param src: the source image, which should be a grayscale or color image.
+    :param ddepth: the desired depth of the destination image, which can be   -1 to use the same depth as the source image, or one of np.uint8, np.uint16, np.int16, np.float32 or np.float64.
+    :param ksize: the size of the kernel used to compute the Laplacian filter, which can be 1, 3, 5 or 7. A larger kernel size will produce more accurate results, but also more noise.
+    :param scale: an optional scaling factor applied to the computed Laplacian values, which can be used to enhance or reduce the effect of the filter.
+    :[aram delta: an optional value added to the computed Laplacian values, which can be used to shift the output image intensity range.
+    :param borderType: an optional flag that specifies how to handle the image borders, which can be one of 'default', 'replicate', 'reflect', or 'constant'.
+
+    """
+
+    # Convert the source image to a numpy array
+    src = np.array(src)
+
+    # Get the shape and depth of the source image
+    height, width = src.shape[:2]
+    src_depth = src.dtype
+
+    # If ddepth is -1, use the same depth as the source image
+    if ddepth == -1:
+        ddepth = src_depth
+
+    # Create a Laplacian kernel matrix according to the ksize
+    if ksize == 1:
+        kernel = np.array([[0, -1, 0], [-1, 4, -1], [0, -1, 0]])
+    elif ksize == 3:
+        kernel = np.array([[0, 1, 0], [1, -4, 1], [0, 1, 0]])
+    elif ksize == 5:
+        kernel = np.array(
+            [
+                [0, 0, -1, 0, 0],
+                [0, -1, -2, -1, 0],
+                [-1, -2, 16, -2, -1],
+                [0, -1, -2, -1, 0],
+                [0, 0, -1, 0, 0],
+            ]
+        )
+    elif ksize == 7:
+        kernel = np.array(
+            [
+                [0, 0, 0, -1, 0, 0, 0],
+                [0, 0, -2, -3, -2, 0, 0],
+                [0, -2, -7, -10, -7, -2, 0],
+                [-1, -3, -10, 68, -10, -3, -1],
+                [0, -2, -7, -10, -7, -2, 0],
+                [0, 0, -2, -3, -2, 0, 0],
+                [0, 0, 0, -1, 0, 0, 0],
+            ]
+        )
+
+    # Apply the Laplacian kernel using convolution
+    laplacian_result = filter2D(
+        src, ddepth, kernel, delta, borderType=BORDER_DEFAULT, anchor=(0, 0)
+    )
+
+    return laplacian_result
+
+
+if __name__ == "__main__":
+    # read original image
+    img = imread(r"digital_image_processing/image_data/lena.jpg")
+
+    # turn image in gray scale value
+    gray = cvtColor(img, COLOR_BGR2GRAY)
+
+    # Applying gaussian filter
+    blur_image = gaussian_filter(gray, 3, sigma=1)
+
+    # Apply multiple Kernel to detect edges
+    laplacian_image = my_laplacian(blur_image, ddepth=CV_64F, ksize=3)
+
+    imshow("Original image", img)
+    imshow("Deteced edges using laplacian filter", laplacian_image)
+
+    waitKey(0)