使用jupyter(IPython)开发opencv

opencv 的默认使用highgui显示图片，
用命令行运行，能够正常显示

cv.namedWindow("Image")
cv.imshow("Image",img)复制代码

用jupyter则有无反应

本文环境

import sys
import cv2
print("python版本:%s"% sys.version)
print("opencv版本:%s"% cv2.__version__)复制代码

输出

python版本:3.5.3 |Continuum Analytics, Inc.| (default, May 15 2017, 10:43:23) [MSC v.1900 64 bit (AMD64)]
opencv版本:3.2.0复制代码

安装opencv

若是使用Anaconda,则打开 Anaconda Prompt,activate python35切换到相应的python环境

pip install --upgrade setuptools
pip install numpy Matplotlib
pip install opencv-python复制代码

参考 Python环境搭建之OpenCV,可是在jupyter中，运行该博文下一段demo代码，无反应

通过一番google，已解决，现整理以下

jupyter显示opencv图片

以lenna 图为例

import cv2
from matplotlib import pyplot as plt

img = cv2.imread('lenna1.png')

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

plt.imshow(show_img)
plt.show()复制代码

参考自 Quickie: Mix up OpenCV and Jupyter (iPython Notebook) 和官方 Using Matplotlib

opencv读取网络图片

%matplotlib inline
import numpy as np
import cv2
from matplotlib import pyplot as plt
import urllib.request as ul

data = None

try:

    data = ul.urlopen('http://www.mupin.it/wp-content/uploads/2012/06/lenna1.png').read()

except Exception as e:

    print("Could not download the image: %s " %( e.message))

else:
    data =  np.fromstring(data, np.uint8)
    img_data =  cv2.imdecode(data, cv2.IMREAD_COLOR )
    img_data = cv2.cvtColor(img_data, cv2.COLOR_BGR2RGB)
    plt.imshow(img_data)
    plt.show()复制代码

本示例用的环境是python:3.5.3 和 opencv:3.2.0，在opencv3.x中已经不存在cv2.CV_LOAD_IMAGE_COLOR,根据 Python OpenCV load image from byte string ，改为cv2.IMREAD_COLOR

大部分代码 参考自 Quickie: Grab an image from the web with urllib2 and OpenCV

OpenCV入门教程

博客 anjia.ml/2017/07/16/…
掘金 juejin.im/post/596b3e…
简书 www.jianshu.com/p/69af8b1dc…