使用OpenCV运行眼动状态检测Demo时遇cv2.imshow未实现错误求助

This error pops up because your OpenCV 3.3.1 installation was compiled without the necessary GUI backend dependencies—specifically, GTK2, which cv2.imshow() relies on to display windows. Let’s walk through two solutions: a permanent fix by rebuilding OpenCV, and a quick workaround if you just need to test your demo right now.

Permanent Fix: Rebuild OpenCV with GTK2 Support #

Follow these steps to recompile OpenCV with the required dependencies:

1. Uninstall your current OpenCV installation

   • If you installed via pip:

     pip uninstall opencv-python
     

   • If you compiled from source, navigate to your OpenCV build directory and run:

     sudo make uninstall
     

2. Install the required development packages
   Update your package list and install the dependencies mentioned in the error:

   sudo apt update && sudo apt install libgtk2.0-dev pkg-config
   

3. Compile and install OpenCV 3.3.1 from source

   • Clone the 3.3.1 branch of OpenCV:

     git clone https://github.com/opencv/opencv.git -b 3.3.1
     cd opencv && mkdir build && cd build
     

   • Run CMake to configure the build, ensuring GTK2 is enabled:

     cmake -D CMAKE_BUILD_TYPE=RELEASE \
           -D CMAKE_INSTALL_PREFIX=/usr/local \
           -D WITH_GTK=ON \
           -D WITH_GTK2=ON ..
     

     Check the CMake output to confirm you see GTK2: YES—this means the dependency was detected correctly.
   • Compile and install:

     make -j$(nproc)  # Uses all available CPU cores for faster compilation
     sudo make install
     sudo ldconfig
     

Quick Workaround: Use Matplotlib to Display Images #

If you don’t want to rebuild OpenCV right now, you can replace cv2.imshow() with Matplotlib’s image display function. Modify your code like this:

import cv2
import matplotlib.pyplot as plt

def image_demo(json_path,weights_path,image_path): 
    img = cv2.imread(image_path)
    # Convert OpenCV's BGR format to Matplotlib's RGB format
    img_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
    plt.imshow(img_rgb)
    plt.title("Image")
    plt.show()

This bypasses OpenCV’s GUI system entirely and uses Matplotlib’s built-in display capabilities.

Why This Error Happens #

cv2.imshow() isn’t a standalone function—it requires a GUI backend library to render windows. When you installed OpenCV 3.3.1, the libgtk2.0-dev package was missing, so the compiler skipped building the GUI module. Rebuilding with the dependency ensures the cvShowImage function (which powers cv2.imshow()) is included in your OpenCV library.

内容的提问来源于stack exchange，提问作者Samuel Mideksa