OpenCV installation
Execute the following commands to install the opencv-python library (core library) and opencv-contrib-python library (contribution library). Note: After the command is copied, it must be executed on a single line, without a new line in the middle.
# 安装opencv核心库
pip3 install --user opencv-python==3.4.2.16 --index-url https://pypi.tuna.tsinghua.edu.cn/simple/ --trusted-host https://pypi.tuna.tsinghua.edu.cn
# 安装opencv贡献库
pip3 install --user opencv-contrib-python==3.4.2.16 --index-url https://pypi.tuna.tsinghua.edu.cn/simple/ --trusted-host https://pypi.tuna.tsinghua.edu.cn
Basic video processing
1) Read the camera
import numpy as np
import cv2
cap = cv2.VideoCapture(0) # 实例化VideoCapture对象, 0表示第一个摄像头
while cap.isOpened():
ret, frame = cap.read() # 捕获帧
cv2.imshow("frame", frame)
c = cv2.waitKey(1) # 等待1毫秒，等待用户输入
if c == 27: # ESC键
break
cap.release() # 释放摄像头
cv2.destroyAllWindows()
2) Play video files
import numpy as np
import cv2
cap = cv2.VideoCapture("D:\\tmp\\min_nong.mp4") # 打开视频文件
while cap.isOpened():
ret, frame = cap.read() # 读取帧
cv2.imshow("frame", frame) # 显示
c = cv2.waitKey(25)
if c == 27: # ESC键
break
cap.release() # 释放视频设备
cv2.destroyAllWindows()
3) Capture and save the video
import numpy as np
import cv2
""" 编解码4字标记值说明
cv2.VideoWriter_fourcc（'I','4','2','0'）表示未压缩的YUV颜色编码格式，色度子采样为4:2:0。
该编码格式具有较好的兼容性，但产生的文件较大，文件扩展名为.avi。
cv2.VideoWriter_fourcc（'P','I','M','I'）表示 MPEG-1编码类型，生成的文件的扩展名为.avi。
cv2.VideoWriter_fourcc（'X','V','I','D'）表示MPEG-4编码类型。如果希望得到的视频大小为平均值，可以选用这个参数组合。
该组合生成的文件的扩展名为.avi。
cv2.VideoWriter_fourcc（'T','H','E','O'）表示Ogg Vorbis编码类型，文件的扩展名为.ogv。
cv2.VideoWriter_fourcc（'F','L','V','I'）表示Flash视频，生成的文件的扩展名为.flv。
"""
cap = cv2.VideoCapture(0)
fourcc = cv2.VideoWriter_fourcc("I", "4", "2", "0") # 编解码4字标记值
out = cv2.VideoWriter("output.avi", # 文件名
fourcc, # 编解码类型
20, # fps(帧速度)
(640, 480)) # 视频分辨率
while cap.isOpened():
ret, frame = cap.read() # 读取帧
if ret == True:
out.write(frame) # 写入帧
cv2.imshow("frame", frame)
if cv2.waitKey(1) == 27: # ESC键
break
else:
break
cap.release()
out.release()
cv2.destroyAllWindows()
Comprehensive case
1) Use OpenCV to achieve image correction
【mission details】
When we analyze and detect the target in the image, the target often has a certain tilt angle, and the images taken under natural conditions are completely flat. Therefore, the process of "righting" the inclined target is called image correction. The original images used in this case are as follows:
# 2021-01-19 星期二 16:10
# 11_img_rectify.py
# 图像形状矫正
import cv2
import math
import numpy as np
# 读取数据
im = cv2.imread('../img_data/paper.jpg')
im_gray = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)
cv2.imshow('im_gray', im_gray)
cv2.imwrite('im_gray.jpg',im_gray)#保存图片
# 二值化
# t,im_bin=cv2.threshold(im_gray,
# 210,255,
# cv2.THRESH_BINARY)
# cv2.imshow('im_bin',im_bin)
# 模糊、膨胀合并过细的细节
im_blur = cv2.GaussianBlur(im_gray, (5, 5), 0)
im_dilate = cv2.dilate(im_blur, (3, 3))
cv2.imshow('im_dilate', im_dilate)
cv2.imwrite('im_dilate.jpg',im_dilate)#保存图片
# canny边沿提取
im_canny = cv2.Canny(im_dilate, # 图像
30, # 滞后阈值
120) # 模糊度
cv2.imshow('im_canny', im_canny)
cv2.imwrite('im_canny.jpg',im_canny)#保存图片
# 在canny边沿提取结果之上查找外部轮廓
img, cnts, hie = cv2.findContours(
im_canny, # canny边沿提取结果
cv2.RETR_EXTERNAL, # 只提取外部轮廓
cv2.CHAIN_APPROX_SIMPLE) # 简化格式保存轮廓
# 绘制轮廓
im_cnt = cv2.drawContours(im, # 原图上绘制
cnts, # 轮廓数据
-1, # 绘制所有轮廓
(0, 0, 255), 2) # 颜色、粗细
cv2.imshow('in_cnt', im_cnt)
cv2.imwrite('in_cnt.jpg',im_cnt)#保存图片
# 计算轮廓面积，由大到小排序
if len(cnts) > 0: # 轮廓数量大于0、
# 排序
cnts = sorted(cnts, # 待排序容器
key=cv2.contourArea, # 排序依据：面积
reverse=True) # 倒序排列
# 对轮廓进行多边形拟合
for cnt in cnts:
peri = cv2.arcLength(cnt, True) # 计算轮廓周长
approx = cv2.approxPolyDP(cnt, # 轮廓
0.02 * peri, # 精度
True) # 产生封闭多边形
if len(approx) == 4: # 四边形
docCnt = approx
break
print(docCnt)
# 用圆圈标记角点
points = []
for peak in docCnt:
peak = peak[0]
cv2.circle(im, # 在原图上绘制
tuple(peak), # 将角点转换为元祖
10, # 圆圈半径大小
(0, 0, 255), 2) # 颜色、粗细
points.append(peak)
cv2.imshow('im_circle', im)
cv2.imwrite('im_circle.jpg',im)#保存图片
# 校正
src = np.float32([points[0], points[1], points[2], points[3]]) # 原来逆时针方向四个点
"""
dst=np.float32([[0,0],[0,488],[337,488],[337,0]])#对应变换后逆时针方向四个点
m=cv2.getPerspectiveTransform(src,dst)#生成透视变换矩阵
result=cv2.warpPerspective(im_gray.copy(),
m,
(337,448))#透视变换
cv2.imshow('result',result)
"""
h = int(math.sqrt((points[1][0] - points[0][0]) ** 2 + (points[1][1]
- points[0][1]) ** 2)) # 宽度
w = int(math.sqrt((points[2][0] - points[1][0]) ** 2 + (points[2][1]
- points[1][1]) ** 2)) # 宽度
print('w:', w, 'h:', h)
dst = np.float32([[0, 0], [0, h], [w, h], [w, 0]])
m = cv2.getPerspectiveTransform(src, dst) # 生成透视变换矩阵
result = cv2.warpPerspective(im_gray.copy(), m, (w, h)) # 透视变换
cv2.imshow('result', result)
cv2.imwrite('result.jpg',result)#保存图片
cv2.waitKey()
cv2.destroyAllWindows()
