深度學(xué)習(xí)數(shù)據(jù)預(yù)處理中常用opencv讀入圖片，一般在__getitem__函數(shù)中調(diào)用。本文主要介紹opencv讀取圖片的一些細(xì)節(jié)以及注意事項。本文中使用的圖片見第6節(jié)“opencv測試使用的圖片”。

1.如何通過opencv顯示圖片？

通過Jupyter Notebook運(yùn)行OpenCV，運(yùn)行cv2.imshow()時報錯。

import cv2

img = cv2.imread(img_path, cv2.IMREAD_UNCHANGED)
cv2.imshow("img", img)

報錯：

---------------------------------------------------------------------------
error                                     Traceback (most recent call last)
Input In [60], in 1>()
----> 1 cv2.imshow("img", img)

error: OpenCV(4.6.0) /io/opencv/modules/highgui/src/window.cpp:1267: error: (-2:Unspecified error) The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Cocoa support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script in function 'cvShowImage'

解決方法：先卸載opencv，再重新安裝。（ 未解決，可直接轉(zhuǎn)下述方法 ）

conda remove opencv
conda install -c menpo opencv

pip install --upgrade pip
pip install opencv-contrib-python

參考鏈接：https://blog.csdn.net/tudou2013goodluck/article/details/108402055

報錯：內(nèi)核崩潰。

Kernel Restarting
The kernel for Untitled.ipynb appears to have died. It will restart automatically.

解決方法 ：opencv讀取圖片，之后使用matplotlib庫顯示圖片。

下述兩個代碼均可以：

%matplotlib inline
from matplotlib import pyplot as plt
import cv2

img = cv2.imread('圖片路徑')
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

plt.imshow(img)
plt.show()

import cv2

def cv2_imshow(a, **kwargs):
    a = a.clip(0, 255).astype('uint8')
    # cv2 stores colors as BGR; convert to RGB
    if a.ndim == 3:
        if a.shape[2] == 4:
            a = cv2.cvtColor(a, cv2.COLOR_BGRA2RGBA)
        else:
            a = cv2.cvtColor(a, cv2.COLOR_BGR2RGB)

    return plt.imshow(a, **kwargs)

參考鏈接：https://github.com/jupyter/notebook/issues/3935

2.opencv讀入圖片默認(rèn)為BGR

img_path = "00001.png"    # 圖片路徑
mask_path = "00001_matte.png"  # 對應(yīng)的圖像分割標(biāo)簽路徑

%matplotlib inline
from matplotlib import pyplot as plt
import cv2
import numpy as np

img_bgr = cv2.imread(img_path, cv2.IMREAD_COLOR)
img_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)  # BGR->RGB

plt.imshow(img)
plt.show()