1 說明：

=====

1.1 完整版：汽車車道動態(tài)視頻檢測講解和注釋版代碼，小白秒懂。

1.2 python+OpenCV+moviepy+numpy為主的技術(shù)要點。

1.3 代碼來源：

https://github.com/linghugoogle/CarND-Advanced-Lane-Lines #雖然感覺也是fork別人的，別忘了給他點個贊star

1.4 感謝原作者，并對文件進(jìn)行修改和代碼進(jìn)行刪減，注釋等操作，便于操作和理解。

1.5 應(yīng)用：無人駕駛汽車技術(shù)，熱門！

2 效果展示：由于gif≤10MB，所以是節(jié)選。

================================

2.1 原視頻節(jié)選：

2.2 處理后視頻節(jié)選：

3 準(zhǔn)備：

=====

3.1 環(huán)境：python3.8+OpenCV4.2.0+deepin-linux操作系統(tǒng)。

3.2 文件結(jié)構(gòu)：

github下載下來

project_video.mp4 ：原始視頻，未進(jìn)行標(biāo)注

vedio_out：文件夾為輸出被標(biāo)注的視頻(處理后的視頻文件夾)

camera_cal：相機(jī)參數(shù)標(biāo)定文件夾。

4 代碼講解：

=========

4.1 line.py代碼：

# -*- coding: utf-8 -*-#導(dǎo)入模塊import numpy as np#定義line這個類class Line(): #初始化參數(shù) def __init__(self): # was the line detected in the last iteration? self.detected = False # x values of the last n fits of the line self.recent_fitted = [np.array([False])] #average x values of the fitted line over the last n iterations self.bestx = None #polynomial coefficients averaged over the last n iterations self.best_fit = None #polynomial coefficients for the most recent fit self.current_fit = [np.array([False])] #radius of curvature of the line in some units self.radius_of_curvature = None #distance in meters of vehicle center from the line self.line_base_pos = None #difference in fit coefficients between last and new fits self.diffs = np.array([0,0,0], dtype='float') #x values for detected line pixels self.allx = None #y values for detected line pixels self.ally = None #檢測偵測 def check_detected(self): if (self.diffs[0] < 0.01 and self.diffs[1] < 10.0 and self.diffs[2] < 1000.) and len(self.recent_fitted) > 0: return True else: return False #更新 def update(self,fit): if fit is not None: if self.best_fit is not None: self.diffs = abs(fit - self.best_fit) if self.check_detected(): self.detected =True if len(self.recent_fitted)>10: self.recent_fitted = self.recent_fitted[1:] self.recent_fitted.append(fit) else: self.recent_fitted.append(fit) self.best_fit = np.average(self.recent_fitted, axis=0) self.current_fit = fit else: self.detected = False else: self.best_fit = fit self.current_fit = fit self.detected=True self.recent_fitted.append(fit)

4.2 utils.py代碼省略。

4.3 main-pipeline.py(就是代碼為：pipeline.py)

# -*- coding: utf-8 -*-#第1步：導(dǎo)入模塊import osimport cv2import matplotlib.pyplot as pltimport numpy as npfrom moviepy.editor import VideoFileClipimport line #自定義模塊import utils #自定義模塊#第2步：圖片閾值處理def thresholding(img): #setting all sorts of thresholds x_thresh = utils.abs_sobel_thresh(img, orient='x', thresh_min=10 ,thresh_max=230) mag_thresh = utils.mag_thresh(img, sobel_kernel=3, mag_thresh=(30, 150)) dir_thresh = utils.dir_threshold(img, sobel_kernel=3, thresh=(0.7, 1.3)) hls_thresh = utils.hls_select(img, thresh=(180, 255)) lab_thresh = utils.lab_select(img, thresh=(155, 200)) luv_thresh = utils.luv_select(img, thresh=(225, 255)) #Thresholding combination threshholded = np.zeros_like(x_thresh) threshholded[((x_thresh == 1) & (mag_thresh == 1)) | ((dir_thresh == 1) & (hls_thresh == 1)) | (lab_thresh == 1) | (luv_thresh == 1)] = 1 return threshholded#第3步：視頻擬合和圖片糾正def processing(img,object_points,img_points,M,Minv,left_line,right_line): #camera calibration, image distortion correction undist = utils.cal_undistort(img,object_points,img_points) #get the thresholded binary image thresholded = thresholding(undist) #perform perspective transform thresholded_wraped = cv2.warpPerspective(thresholded, M, img.shape[1::-1], flags=cv2.INTER_LINEAR) #perform detection if left_line.detected and right_line.detected: left_fit, right_fit, left_lane_inds, right_lane_inds = utils.find_line_by_previous(thresholded_wraped,left_line.current_fit,right_line.current_fit) else: left_fit, right_fit, left_lane_inds, right_lane_inds = utils.find_line(thresholded_wraped) left_line.update(left_fit) right_line.update(right_fit) #draw the detected laneline and the information area_img = utils.draw_area(undist,thresholded_wraped,Minv,left_fit, right_fit) curvature,pos_from_center = utils.calculate_curv_and_pos(thresholded_wraped,left_fit, right_fit) result = utils.draw_values(area_img,curvature,pos_from_center) return result#第4步：步驟：劃線-校正-讀取原視頻和生成修改后的視頻#劃線left_line = line.Line() #左線right_line = line.Line() #右線#獲取棋盤格圖片#使用提供的一組棋盤格圖片計算相機(jī)校正矩陣(camera calibration matrix)和失真系數(shù)(distortion coefficients).cal_imgs = utils.get_images_by_dir('/home/xgj/Desktop/v-carline-good/camera_cal')#計算object_points,img_pointsobject_points,img_points = utils.calibrate(cal_imgs,grid=(9,6))M,Minv = utils.get_M_Minv()#需要修改的視頻：原視頻project_video_clip = VideoFileClip("/home/xgj/Desktop/v-carline-good/project_video.mp4")#輸出修改后的視頻：完成視頻project_outpath = '/home/xgj/Desktop/v-carline-good/vedio_out/project_video_out.mp4'#制作視頻project_video_out_clip = project_video_clip.fl_image(lambda clip: processing(clip,object_points,img_points,M,Minv,left_line,right_line))project_video_out_clip.write_videofile(project_outpath, audio=False)

5 完結(jié)：

=====

5.1 以上代碼完整，但制作視頻估計花20分鐘，我也是將代碼最簡化跑起來。

5.2 如果逐步深入分析，可能要從基本開始。

5.3 可以參考這篇文章：

https://zhuanlan.zhihu.com/p/46146266

總結(jié)

以上是生活随笔為你收集整理的python 实现显著性检测_强！汽车车道视频检测：python+OpenCV为主实现的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯，歡迎將生活随笔推薦給好友。