imWeightedThresholdedheq

This module attempts to enhance contrast of a given image or video by employing a method called weighted thresholded histogram equalization (WTHE). This method seeks to improve on preformance of the conventional histogram equalization method by adding controllable parameters to it. By weighting and thresholding the PMF of the image before performing histogram equalization, two parameters are introduced that can be changed manually, but by experimenting on a variety of images, optimal values for both parameters are calculated (r = 0.5, v = 0.5).

You can access the article that came up with this method here.

Installation

Run the following to install:

pip install imWeightedThresholdedheq

Usage

For images

imWeightedThresholdedheq --input 'Plane.jpg' --output 'Plane-imWeightedThresholdedheq.jpg'

For videos

vid2dhisteq --input 'assets/Arctic-Convoy-With-Giant-Mack-Trucks.mp4' --output 'assets/Arctic-Convoy-With-Giant-Mack-Trucks-imWeightedThresholdedheq.mp4'

Or

import numpy as np
import cv2
from imWeightedThresholdedheq import imWTHeq

cap = cv2.VideoCapture('assets/Arctic-Convoy-With-Giant-Mack-Trucks.mp4')

# output video without sound
video_out_name = 'assets/Arctic-Convoy-With-Giant-Mack-Trucks-imWeightedThresholdedheq.mp4'
i = 0
j = 0
Wout_list = np.zeros((10))
while(cap.isOpened()):
    ret, frame = cap.read()
    if ret == False:
        break
    frame_hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
    frame_v   = frame_hsv[:, :, 2].copy()
    image_heq, Wout = imWTHeq(frame_v, Wout_list, r=0.5, v=0.5)
    Wout_list[j] = Wout
    j += 1
    if j == 10:
        j = 0
    frame_hsv[:, :, 2] = image_heq
    frame_eq = cv2.cvtColor(frame_hsv, cv2.COLOR_HSV2BGR)

    fps = cap.get(cv2.CAP_PROP_FPS)
    if i==0:
        h, w, d = frame_eq.shape
        fourcc = cv2.VideoWriter_fourcc(*'mp4v')
        video_out = cv2.VideoWriter(video_out_name, fourcc, fps, (w, h))
    video_out.write(frame_eq)

    i+=1
cv2.destroyAllWindows()
video_out.release()

Showcase

• A 5 minutes comparative video: https://youtu.be/5H_EY_ugmzg
• A sample video and its enhanced version by WTHE method