#pip install opencv-python==4.4.0.42 opencv-contrib-python==4.4.0.42
 
import cv2
import numpy as np
from matplotlib import pyplot as plt
 
import matplotlib.image as mpimg
import matplotlib as mpl
from PIL import Image
mpl.rcParams['figure.dpi'] = 200
 
#加載圖像
img = cv2.imread('test.png')
plt.axis('off')
plt.imshow(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
plt.show()
#灰度圖像
img1 = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
#cornerHarris要求img是flaot32位
img1 = np.float32(img1)
#Harris角點(diǎn)檢測函數(shù)
#　　? img - 數(shù)據(jù)類型為 float32 的輸入圖像。
#　　? blockSize - 角點(diǎn)檢測中要考慮的領(lǐng)域大小。
#　　? ksize - Sobel 求導(dǎo)中使用的窗口大小
#　　? k - Harris 角點(diǎn)檢測方程中的自由參數(shù),取值參數(shù)為 [0,04,0.06].
dst = cv2.cornerHarris(img1,10,3,0.04)
plt.axis('off')
plt.imshow(dst, cmap='gray')
plt.show()
#膨脹
dst = cv2.dilate(dst,None)
plt.axis('off')
plt.imshow(dst, cmap='gray')
plt.show()
#顯示經(jīng)過處理后的圖片
threshold = 0.01*dst.max()
img[dst>threshold]=[255,0,0] 
#[255,0,0] - 點(diǎn)的顏色：藍(lán)色, [0,255,0] - 綠色, [0,0,255] - 紅色,
#[0,0,0] -白色, [255,255,255] - 黑色
plt.axis('off')
plt.imshow(img, cmap='gray')
plt.show()
 
#使用另一種角點(diǎn)檢測函數(shù)
corners = cv2.goodFeaturesToTrack(gray, 6, 0.01, 5)
corners = np.int0(corners) 
for i in corners:
    x, y = i.ravel()
    cv2.circle(img, (x, y), 5, [0,0,0], -1) 
plt.axis('off')
plt.imshow(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
plt.show()
 
#直線檢測
 
img = cv2.imread("road.jpeg")
plt.axis('off')
plt.imshow(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
plt.show()
 
#canny 方法檢測邊緣 返回二值圖像
edges = cv2.Canny(gray, 150, 300)
plt.axis('off')
plt.imshow(edges, cmap='gray')
plt.show()
#HoughLinesP方法判斷哪些邊緣是直線
 
lines = cv2.HoughLinesP(edges, rho=1.0,theta=np.pi/180,threshold=20,minLineLength=30,maxLineGap=10)
line_img = np.zeros((img.shape[0], img.shape[1], 3), dtype=np.uint8) 
line_color = [0, 255, 0]
line_thickness = 2
dot_color = [0, 255, 0]
dot_size = 3
 
#講檢測的直線疊加到原圖
for line in lines:
    for x1, y1, x2, y2 in line:
        cv2.line(line_img, (x1, y1), (x2, y2), line_color, line_thickness)
        cv2.circle(line_img, (x1, y1), dot_size, dot_color, -1)
        cv2.circle(line_img, (x2, y2), dot_size, dot_color, -1)
final = cv2.addWeighted(img, 0.8, line_img, 1.0, 0.0) 
plt.axis('off')
plt.imshow(cv2.cvtColor(final, cv2.COLOR_BGR2RGB))
plt.show()
 

#直線檢測
#使用霍夫直線變換做直線檢測，前提條件：邊緣檢測已經(jīng)完成
import cv2 as cv
import numpy as np
import matplotlib.pylab as plt
 
#標(biāo)準(zhǔn)霍夫線變換
def line_detection(image):
    gray = cv.cvtColor(image, cv.COLOR_RGB2GRAY)
    edges = cv.Canny(gray, 50, 150, apertureSize=3)  #apertureSize參數(shù)默認(rèn)其實(shí)就是3
    cv.imshow("edges", edges)
    lines = cv.HoughLines(edges, 1, np.pi/180, 80)
    for line in lines:
        rho, theta = line[0]  #line[0]存儲的是點(diǎn)到直線的極徑和極角，其中極角是弧度表示的。
        a = np.cos(theta)   #theta是弧度
        b = np.sin(theta)
        x0 = a * rho    #代表x = r * cos（theta）
        y0 = b * rho    #代表y = r * sin（theta）
        x1 = int(x0 + 1000 * (-b)) #計算直線起點(diǎn)橫坐標(biāo)
        y1 = int(y0 + 1000 * a)    #計算起始起點(diǎn)縱坐標(biāo)
        x2 = int(x0 - 1000 * (-b)) #計算直線終點(diǎn)橫坐標(biāo)
        y2 = int(y0 - 1000 * a)    #計算直線終點(diǎn)縱坐標(biāo)    注：這里的數(shù)值1000給出了畫出的線段長度范圍大小，數(shù)值越小，畫出的線段越短，數(shù)值越大，畫出的線段越長
        cv.line(image, (x1, y1), (x2, y2), (0, 0, 255), 2)    #點(diǎn)的坐標(biāo)必須是元組，不能是列表。
    cv.imshow("image-lines", image)
 
#統(tǒng)計概率霍夫線變換
def line_detect_possible_demo(image):
    gray = cv.cvtColor(image, cv.COLOR_RGB2GRAY)
    edges = cv.Canny(gray, 50, 150, apertureSize=3)  # apertureSize參數(shù)默認(rèn)其實(shí)就是3
    lines = cv.HoughLinesP(edges, 1, np.pi / 180, 60, minLineLength=60, maxLineGap=5)
    for line in lines:
        x1, y1, x2, y2 = line[0]
        cv.line(image, (x1, y1), (x2, y2), (0, 0, 255), 2)
    cv.imshow("line_detect_possible_demo",image)
 
src = cv.imread(r'..\edge.jpg')
print(src.shape)
cv.namedWindow('input_image', cv.WINDOW_AUTOSIZE)
cv.imshow('input_image', src)
line_detection(src)
src = cv.imread(r'..\edge.jpg') #調(diào)用上一個函數(shù)后，會把傳入的src數(shù)組改變，所以調(diào)用下一個函數(shù)時，要重新讀取圖片
line_detect_possible_demo(src)
cv.waitKey(0)
cv.destroyAllWindows()