import cv2

import numpy as np

def adjust_gamma(image, gamma=1.0):

invGamma = 1.0 / gamma

table = np.array([((i / 255.0) ** invGamma) * 255 for i in np.arange(0, 256)]).astype("uint8")

return cv2.LUT(image, table)

def lab_contrast(image):

return cv2.cvtColor(limg, cv2.COLOR_LAB2BGR)

cap = cv2.VideoCapture(0)

scale = 3

cap.set(cv2.CAP_PROP_FRAME_WIDTH, int(scale * cap.get(cv2.CAP_PROP_FRAME_WIDTH)))

cap.set(cv2.CAP_PROP_FRAME_HEIGHT, int(scale * cap.get(cv2.CAP_PROP_FRAME_HEIGHT)))

knn = cv2.createBackgroundSubtractorKNN(detectShadows=True)

first_frame = None

while True:

ret, frame = cap.read()

y1 = int(frame.shape[0] * .05)

x1 = int(frame.shape[1] * .05)

y2 = int(frame.shape[0] * .95)

x2 = int(frame.shape[1] * .5)

left_up = (x1, y1)

right_down = (x2, y2)

# gray = cv2.GaussianBlur(cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY), (21, 21), 0)[y1:y2, x1:x2]

cut = frame[y1:y2, x1:x2]

# gamma = cv2.medianBlur(adjust_gamma(cut, 1), 7)

# contrast = cv2.medianBlur(lab_contrast(gamma), 7)

gray = cv2.medianBlur(cv2.cvtColor(cut, cv2.COLOR_BGR2GRAY), 11)

if first_frame is None:

first_frame = gray

continue

# delta = knn.apply(gray)

cv2.rectangle(frame, left_up, right_down, (0, 255, 0), 5)

cv2.imshow('frame', frame)

cv2.moveWindow('frame', 0, 0)

delta = cv2.absdiff(first_frame, gray)

cv2.imshow('delta', delta)

cv2.moveWindow('delta', 1300, 0)

_, thresh = cv2.threshold(delta, 50, 255, cv2.THRESH_BINARY)

thresh = cv2.dilate(thresh, None, iterations=3)

cv2.imshow('thresh', thresh)

cv2.moveWindow('thresh', 1920, 0)

if cv2.waitKey(1) & 0xFF == ord('q'):

break

cap.release()

cv2.destroyAllWindows()