Lane Line Detection Using Canny and HoughLinesP

Code :

import matplotlib.pylab as plt

import cv2

import NumPy as np

def region_of_interest(img, vertices):

    mask = np.zeros_like(img)

    #channel_count = img.shape[2]

    match_mask_color = 255

    cv2.fillPoly(mask, vertices, match_mask_color)

    masked_image = cv2.bitwise_and(img, mask)

    return masked_image

def drow_the_lines(img, lines):

    img = np.copy(img)

    blank_image = np.zeros((img.shape[0], img.shape[1], 3), dtype=np.uint8)

    for line in lines:

        for x1, y1, x2, y2 in line:

            cv2.line(blank_image, (x1,y1), (x2,y2), (0, 0, 255), thickness=4)

    img = cv2.addWeighted(img, 0.8, blank_image, 1, 0.0)

    return img

# = cv2.imread('road.jpg')

#image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)

def process(image):

    print(image.shape)

    height = image.shape[0]

    width = image.shape[1]

    region_of_interest_vertices = [

        (0, height),

        (width/2, height/2),

        (width, height)

    ]

    gray_image = cv2.cvtColor(image, cv2.COLOR_RGB2GRAY)

    canny_image = cv2.Canny(gray_image, 180, 240)

    cropped_image = region_of_interest(canny_image,

                    np.array([region_of_interest_vertices], np.int32),)

    lines = cv2.HoughLinesP(cropped_image,

                            rho=2,

                            theta=np.pi/180,

                            threshold=90,

                            lines=np.array([]),

                            minLineLength=90,

                            maxLineGap=5)

    image_with_lines = drow_the_lines(image, lines)

    return image_with_lines

cap = cv2.VideoCapture('road.mp4')

while cap.isOpened():

    ret, frame = cap.read()

    frame = process(frame)

    cv2.imshow('frame', frame)

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

        break

cap.release()

cv2.destroyAllWindows()

GitHub