Python – 图像处理 – OpenCV常用操作

安装

pip install opencv-python
pip install numpy

pip install opencv-python

pip install numpy

获取图片

生成空白图片

black = np.zeros((300, 300, 3), np.uint8)
black.fill(0) # 黑色

black = np.zeros((300, 300, 3), np.uint8)

black.fill(0) # 黑色

从图片文件读取

img = cv2.imread('1.jpg', cv2.IMREAD_COLOR) # 彩色
img = cv2.imread('1.jpg', 0) # 灰度

img = cv2.imread('1.jpg', cv2.IMREAD_COLOR) # 彩色

img = cv2.imread('1.jpg', 0) # 灰度

从视频文件读取

cap = cv2.VideoCapture('1.mp4')
if not cap.isOpened():
    print("video cannot open!\n")

while True:
    ret, img = cap.read()
    if not ret:
        break

cap = cv2.VideoCapture('1.mp4')

if not cap.isOpened():

print("video cannot open!\n")

while True:

ret, img = cap.read()

if not ret:

break

从在线视频流读取

cap = cv2.VideoCapture('rtmp://****.*****/*****')
if not cap.isOpened():
    print("video cannot open!\n")

while True:
    ret, img = cap.read()
    if not ret:
        break

cap = cv2.VideoCapture('rtmp://****.*****/*****')

if not cap.isOpened():

print("video cannot open!\n")

while True:

ret, img = cap.read()

if not ret:

break

读取和调整图片大小

获取图片大小

h, w, _ = img.shape

1 2	h, w, _ = img.shape

调整图片大小

img = cv2.resize(img, (500, 500))

1 2	img = cv2.resize(img, (500, 500))

常用绘图操作

二值化

ret, mask = cv2.threshold(img, 200, 255, cv2.THRESH_BINARY)
# 大于200的置为255

ret, mask = cv2.threshold(img, 200, 255, cv2.THRESH_BINARY)

# 大于200的置为255

绘制文字

# 图片 文字 左上角坐标 字体 大小 颜色 粗细
cv2.putText(img, "txt", (x, y), 0, 2, (255, 255, 255), 3)

# 图片文字左上角坐标字体大小颜色粗细

cv2.putText(img, "txt", (x, y), 0, 2, (255, 255, 255), 3)

绘制中文文本

from PIL import Image, ImageDraw, ImageFont

def cv2ImgAddText(img, text, left, top, textColor=(0, 255, 0), textSize=20):
    if (isinstance(img, np.ndarray)):  
        img = Image.fromarray(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))

    draw = ImageDraw.Draw(img)
    # 字体
    fontStyle = ImageFont.truetype(
        "MSYH.TTC", textSize, encoding="utf-8")

    # 绘制文本
    draw.text((left, top), text, textColor, font=fontStyle)

    return cv2.cvtColor(np.asarray(img), cv2.COLOR_RGB2BGR)

image = cv2ImgAddText(image, '中文', 5, 5, (0, 255, 0), textSize=20)

from PIL import Image, ImageDraw, ImageFont

def cv2ImgAddText(img, text, left, top, textColor=(0, 255, 0), textSize=20):

if (isinstance(img, np.ndarray)):

img = Image.fromarray(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))

draw = ImageDraw.Draw(img)

# 字体

fontStyle = ImageFont.truetype(

"MSYH.TTC", textSize, encoding="utf-8")

# 绘制文本

draw.text((left, top), text, textColor, font=fontStyle)

return cv2.cvtColor(np.asarray(img), cv2.COLOR_RGB2BGR)

image = cv2ImgAddText(image, '中文', 5, 5, (0, 255, 0), textSize=20)

绘制矩形

x1, y1 = 20, 20
x2, y2 = 100, 100
# 图片 左上角坐标 右下角坐标 颜色 粗细
cv2.rectangle(img, (x1, y1), (x2, y2), (255, 255, 255), 3)

x1, y1 = 20, 20

x2, y2 = 100, 100

# 图片左上角坐标右下角坐标颜色粗细

cv2.rectangle(img, (x1, y1), (x2, y2), (255, 255, 255), 3)

绘制直线

cv2.line(img, (x1, y1), (x2, y2), Scalar(33, 33, 133), 2)

1 2	cv2.line(img, (x1, y1), (x2, y2), Scalar(33, 33, 133), 2)

绘制多边形

poly = np.array([[150,50], [140,140], [200,170], [250,250], [150,50]], np.int32)
cv2.polylines(img, [poly], True, (0,255,255), 3)

poly = np.array([[150,50], [140,140], [200,170], [250,250], [150,50]], np.int32)

cv2.polylines(img, [poly], True, (0,255,255), 3)

颜色转换

RGB转为BGR

img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

1 2	img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

将图片的一部分复制到另一张图片

src = img[pt1y:pt2y, pt1x:pt2x]
black[pt1y:pt2y, pt1x:pt2x] = src

src = img[pt1y:pt2y, pt1x:pt2x]

black[pt1y:pt2y, pt1x:pt2x] = src

显示和保存图片

打开一个新窗口显示图片

cv2.imshow('image', img) # 在image窗口显示图片
cv2.imwrite('2.jpg', img) # 保存图片

# 全屏显示
window_name = 'result'
cv2.namedWindow(window_name, cv2.WND_PROP_FULLSCREEN)
cv2.moveWindow(window_name, 0, 0)
cv2.setWindowProperty(window_name, cv2.WND_PROP_FULLSCREEN, cv2.WINDOW_FULLSCREEN)
cv2.imshow("result", img)

cv2.imshow('image', img) # 在image窗口显示图片

cv2.imwrite('2.jpg', img) # 保存图片

# 全屏显示

window_name = 'result'

cv2.namedWindow(window_name, cv2.WND_PROP_FULLSCREEN)

cv2.moveWindow(window_name, 0, 0)

cv2.setWindowProperty(window_name, cv2.WND_PROP_FULLSCREEN, cv2.WINDOW_FULLSCREEN)

cv2.imshow("result", img)

在notebook中显示图片

import cv2
from PIL import Image
import matplotlib.pyplot as plt
import matplotlib
%matplotlib inline

img = cv2.imread('00001.png')

fig, ax = plt.subplots(figsize=(16, 10))
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
ax.imshow(img)
ax.set_axis_off()
plt.show()

import cv2

from PIL import Image

import matplotlib.pyplot as plt

import matplotlib

%matplotlib inline

img = cv2.imread('00001.png')

fig, ax = plt.subplots(figsize=(16, 10))

img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

ax.imshow(img)

ax.set_axis_off()

plt.show()

按任意键继续

if cv2.waitKey(0) & 0xFF == ord('q'): #按下q键
    #cap.release() # 释放视频
    cv2.destroyAllWindows()

if cv2.waitKey(0) & 0xFF == ord('q'): #按下q键

#cap.release() # 释放视频

cv2.destroyAllWindows()

参考

https://blog.csdn.net/what_lei/article/details/49159655
边缘检测
https://blog.csdn.net/weixin_43309755/article/details/89040894

PIL Image库的用法
https://www.cnblogs.com/lyrichu/p/9124504.html