Added histogram equalization to computer vision

computer_vision/histogram_equalization.py

@@ -0,0 +1,33 @@
+import cv2
+import numpy as np
+
+"""
+Histogram Equalization for Image Enhancement
+https://en.wikipedia.org/wiki/Histogram_equalization
+"""
+
+
+def hist_equalization(image):
+    """
+    Returns the histogram equalization image
+    :param image: input image
+    """
+    l_channel, a_channel, b_channel = cv2.split(image)
+    histogram = cv2.calcHist([l_channel], [0], None, [256], [0, 256])
+    histogram_sum = np.sum(histogram)
+    probability_density_function = histogram / histogram_sum
+    cumulative_distribution_function = np.cumsum(probability_density_function)
+    lookup_table = np.round(cumulative_distribution_function * 255).astype(np.uint8)
+    equalized_l = cv2.LUT(l_channel, lookup_table)
+    new_image = cv2.merge((equalized_l, a_channel, b_channel))
+    new_image = cv2.cvtColor(new_image, cv2.COLOR_LAB2BGR)
+    return new_image
+
+
+if __name__ == "__main__":
+    image = cv2.imread("path_to_image")
+    image = cv2.cvtColor(image, cv2.COLOR_BGR2LAB)
+    new_image = hist_equalization(image)
+    cv2.imshow("image", new_image)
+    cv2.waitKey(0)
+    cv2.destroyAllWindows()