open cv

opencv2

        opencv库包含许多计算机视觉的相关库文件，可以使用调用函数来实现计算机视觉的相应功能。大部分的时间都是用来实现对图片，视频的操作，当然在最近的一个news中，其实也不算新闻了，，介绍“刷脸签到应用"（ https://www.leiphone.com/news/201710/Niuqjv1wsqvIc46f.html)最后我也尝试性的实现这个System!不过在轻易实现之前，We 需要做跟多的基础工作，Knowledge 是必要的

        

    1.1 对图片的操作

# -*- coding: cp936 -*-

import cv2

import numpy as np

from matplotlib import pyplot as plt

"this class is used for testing the opencv2 about operating the picture"

class testclass(object):

    "the first test:how to open/write a picture"

    def __init__(self):

        pass

    def dopicture(self,path):

        self.path = path

        img = cv2.imread(self.path,0) #0是一个参数：cv2.IMREAD_COLOR 灰度值

        cv2.namedWindow('picture',cv2.WINDOW_NORMAL) #命名一个图片的窗口

        cv2.imshow('picture',img)  #展示图片

        cv2.imwrite('D:/spicture.png',img)

        cv2.waitKey(0)

        cv2.destroyAllWindows()   #摧毁所有打开的窗口

        print "完成对图片的读取与写入功能"

    "the second test :how to plot a picture by using matplotlib"

    def plotpicture(self,path):

        self.path = path

        img = cv2.imread(self.path,0)

        plt.imshow(img,cmap='gray',interpolation='bicubic')

        plt.xticks([]),plt.yticks([])

        plt.show

def test():

    path = str(raw_input("请输入图片路径：>\n"))

    pic = testclass()

    pic.plotpicture(path)

    print "test is effective!"

test()

    1.2 对视频的处理

        实际上在处理视频的应用中，常会用到两个functions：一个cv2.VideoCapture(),一个cv2.VideoWrite().

import numpy as np

import cv2

cap = cv2.VideoCapture('D:/test.mp4')   #display a video

                  #创建一个Videocapture对象，后面的函数主要用来就是操作这个对象

while(True):

    ret ,frame = cap.read()

    ret = cap.set(3,320)

    gray = cv2.cvtColor(frame,cv2.COLOR_BGR2GRAY)

    cv2.imshow('frame',gray)

    if cv2.waitKey(1) & 0xFF == ord('q'):

        break

cap.release()

cv2.destoryAllWindows()

        知道如何读取视频之后，现在该思考怎么把视频存放在相应的文件夹，主要是Get out输出对象的操作，很大程度上与cap相似

# -*- coding: cp936 -*-

import numpy as np

import cv2

cap = cv2.VideoCapture(0)

fourcc = cv2.cv.CV_FOURCC(*'MJPG')

        #there is a little problem!

        #python3.0+版本采用语句：fourcc = cv2.VideoWriter_fourcc(*'XVID')

out = cv2.VideoWriter('D:/output.avi',fourcc,20.0,(640,480))

while(cap.isOpened()):

    ret ,frame = cap.read()

    ret = cap.set(3,320)

    if(ret == True):

        frame = cv2.flip(frame,0)

        out.write(frame)

        cv2.imshow('frame',frame)

        if cv2.waitKey(1) & 0xFF == ord('q'):

            break

    else:

        break

cap.release()

out.release()

cv2.destroyAllWindows()

前面两个小点只是花生米，开始才刚刚开始，对于一个特征物体，我们更需要的是了解这个特征物的feature，顺带的去get这个feature detection,顺带去添加一些描述description，特征是个什么东东如官方API中一句话“   we may find it difficult to express it in words, but we know what are they ”