1.获取图像的高、宽、通道数

height = image.shape[0]
width = image.shape[1]
channels = image.shape[2]
# 也可以直接写成：
height, width, channels = image.shape[:]

2.计算程序的执行时间-cv2.getTickCount()、cv2.getTickFrequency()

• cv2.getTickCount()  作用:用于返回从操作系统启动到当前所经的计时周期数.
• cv2.getTickFrequency() 作用：用于返回CPU的频率,单位是s.
• 得到某段程序的执行时间: time = (t2 - t1)/cv.getTickFrequency() ,单位是s. t1、t2:分别表示函数执行前和执行后的时间。

3. 图像的创建和修改

img = np.zeros([400, 400, 3], np.uint8)          # 创建多维数组,每个值都是0,类型为uint8.
img[:, :, 0] = np.ones([400, 400]) * 255         # 对B通道的进行幅值,全为255.
cv.imshow("new_image", img)

4. 将图像颜色颠倒-cv2.bitwise_not()

5. 完整代码:

import cv2 as cv
import numpy as np


def access_pixels(image):  # 获取图片高宽通道
    print(image.shape)
    height = image.shape[0]
    width = image.shape[1]
    channels = image.shape[2]
    # 也可以直接写成：
    # height, width, channels = image.shape[:]
    print("width: %s, height: %s, channels: %s"%(width, height, channels))

    new_image = image.copy()  # 复制image图片
    for row in range(height):
        for col in range(width):
            for c in range(channels):
                pv = image[row, col, c]
                new_image[row, col, c] = 256 - pv  # 图像反转

    cv.imshow('pixels_demo', new_image)


def create_image():

    img = np.zeros([400, 400, 3], np.uint8)          # 创建多维数组,每个值都是0,类型uint8.
    img[:, :, 0] = np.ones([400, 400]) * 255         # 对B通道的进行赋值,全为255.
    cv.imshow("new_image", img)

    # 对单通道进行操作
    # img = np.ones([400, 400, 1], np.uint8)
    # img = img * 0
    # cv.imshow("new_image", img)

def inverse(image):
    dst = cv.bitwise_not(image)  # 按位取反，白变黑，黑变白，效果和new_image一样
    cv.imshow("inverse_demo",dst)


if __name__ == '__main__':

    src = cv.imread("../images/CrystalLiu1.jpg")  # 读入图片放进src中
    cv.namedWindow("Crystal Liu", cv.WINDOW_AUTOSIZE)  # 创建窗口, 窗口尺寸自动调整
    cv.imshow("Crystal Liu", src)  # 将src图片放入该创建的窗口中
    create_image()

    t1 = cv.getTickCount()  # 获取当前电脑时钟滴答数

    access_pixels(src)
    inverse(src)
    t2 = cv.getTickCount()
    time = (t2 - t1)/cv.getTickFrequency()  # 函数执行前后滴答数之差与滴答频率之比为前后时间差
    print("time: %s ms"%(time * 1000))
    cv.waitKey(0)