我试图用moviepy运行此车道检测并且它可以工作，但是我需要使用opencv（cv2.VideoCapture（））运行它，但是cv2.imshow只显示输入视频，根本不显示行。 / p>
带有moviepy的代码：
import numpy as np
import cv2
from numpy.polynomial import Polynomial as P
from moviepy.editor import VideoFileClip
import pygame
PREV_LEFT_X1 = None
PREV_LEFT_X2 = None
PREV_RIGHT_X1 = None
PREV_RIGHT_X2 = None
BASE_IMG = None
CANNY_IMG = None
def to_hsv(img):
return cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
def grayscale(img):
return cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
def canny(img, low_threshold, high_threshold):
return cv2.Canny(img, low_threshold, high_threshold)
def gaussian_blur(img, kernel_size):
return cv2.GaussianBlur(img, (kernel_size, kernel_size), 0)
def filter_color(img):
yellow_min = np.array([65, 80, 80], np.uint8)
yellow_max = np.array([105, 255, 255], np.uint8)
yellow_mask = cv2.inRange(img, yellow_min, yellow_max)
white_min = np.array([0, 0, 200], np.uint8)
white_max = np.array([255, 80, 255], np.uint8)
white_mask = cv2.inRange(img, white_min, white_max)
img = cv2.bitwise_and(img, img, mask=cv2.bitwise_or(yellow_mask, white_mask))
return img
def region_of_interest(img, vertices):
#defining a blank mask to start with
mask = np.zeros_like(img)
#defining a 3 channel or 1 channel color to fill the mask with depending on the input image
if len(img.shape) > 2:
channel_count = img.shape[2] # i.e. 3 or 4 depending on your image
ignore_mask_color = (255,) * channel_count
else:
ignore_mask_color = 255
#filling pixels inside the polygon defined by "vertices" with the fill color
cv2.fillPoly(mask, vertices, ignore_mask_color)
#returning the image only where mask pixels are nonzero
masked_image = cv2.bitwise_and(img, mask)
return masked_image
def slope(line):
return (float(line[3]) - line[1]) / (float(line[2]) - line[0])
def draw_lines(img, lines, color=[255, 0, 0], thickness=7):
global PREV_LEFT_X1, PREV_LEFT_X2, PREV_RIGHT_X1, PREV_RIGHT_X2
left_x = []
left_y = []
right_x = []
right_y = []
if lines is not None:
for line in lines:
line = line[0]
s = slope(line)
if 0.3 > s > -0.3:
continue
if s < 0:
if line[0] > img.shape[1] / 2 + 40:
continue
left_x += [line[0], line[2]]
left_y += [line[1], line[3]]
# cv2.line(img, (int(line[0]), int(line[1])), (int(line[2]), int(line[3])), [0, 0, 255], thickness)
else:
if line[0] < img.shape[1] / 2 - 40:
continue
right_x += [line[0], line[2]]
right_y += [line[1], line[3]]
# cv2.line(img, (int(line[0]), int(line[1])), (int(line[2]), int(line[3])), [255, 255, 0], thickness)
y1 = img.shape[0]
y2 = img.shape[0] / 2 + 90
if len(left_x) <= 1 or len(right_x) <= 1:
if PREV_LEFT_X1 is not None:
cv2.line(img, (int(PREV_LEFT_X1), int(y1)), (int(PREV_LEFT_X2), int(y2)), color, thickness)
cv2.line(img, (int(PREV_LEFT_X2), int(y1)), (int(PREV_RIGHT_X2), int(y2)), color, thickness)
return
left_poly = P.fit(np.array(left_x), np.array(left_y), 1)
right_poly = P.fit(np.array(right_x), np.array(right_y), 1)
left_x1 = (left_poly - y1).roots()
right_x1 = (right_poly - y1).roots()
left_x2 = (left_poly - y2).roots()
right_x2 = (right_poly - y2).roots()
if PREV_LEFT_X1 is not None:
left_x1 = PREV_LEFT_X1 * 0.7 + left_x1 * 0.3
left_x2 = PREV_LEFT_X2 * 0.7 + left_x2 * 0.3
right_x1 = PREV_RIGHT_X1 * 0.7 + right_x1 * 0.3
right_x2 = PREV_RIGHT_X2 * 0.7 + right_x2 * 0.3
PREV_LEFT_X1 = left_x1
PREV_LEFT_X2 = left_x2
PREV_RIGHT_X1 = right_x1
PREV_RIGHT_X2 = right_x2
cv2.line(img, (int(left_x1), int(y1)), (int(left_x2), int(y2)), color, thickness)
cv2.line(img, (int(right_x1), int(y1)), (int(right_x2), int(y2)), color, thickness)
def hough_lines(img, rho, theta, threshold, min_line_len, max_line_gap):
lines = cv2.HoughLinesP(img, rho, theta, threshold, np.array([]), minLineLength=min_line_len, maxLineGap=max_line_gap)
line_img = np.zeros((*img.shape, 3), dtype=np.uint8)
draw_lines(line_img, lines)
return line_img
# Python 3 has support for cool math symbols.
def weighted_img(img, initial_img, α=1., β=1., λ=0.):
return cv2.addWeighted(initial_img, α, img, β, λ)
def process_image(base_img):
global BASE_IMG, CANNY_IMG
BASE_IMG = base_img
ysize = base_img.shape[0]
xsize = base_img.shape[1]
image_hsv = to_hsv(base_img)
image_blur = gaussian_blur(image_hsv, 3)
image_color = filter_color(image_blur)
image_canny = canny(image_color, 30, 130)
CANNY_IMG = image_canny
image_interest = region_of_interest(
image_canny,
np.array(
[[(40, ysize), (xsize / 2, ysize / 2 + 40), (xsize / 2, ysize / 2 + 40), (xsize - 40, ysize)]],
dtype=np.int32
)
)
image = hough_lines(image_interest, 1, np.pi / 180, 50, 15, 10)
# return image
return weighted_img(image, base_img, β=250.)
clip = VideoFileClip("25.mp4")
lane_line = clip.fl_image(process_image)
lane_line.preview(audio=False)
pygame.quit()
然后我尝试将opencv与相同的代码一起使用，但是图像结果没有红线，它只是显示原始图像。下面的代码：
cap = cv2.VideoCapture('25.mp4')
success = True
while success:
success, frame = cap.read()
if success == True:
output = process_image(frame)
cv2.imshow('result', output)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
else:
break
cap.release()
cv2.destroyAllWindows()
我做错了什么？