虚拟环境中pip安装OpenCV时出现TLS协议版本错误求助

Hey there, let's work through this install error you're seeing. The root of the problem is two-fold:

• Your pip version is outdated, so it doesn't support the newer TLS protocols that PyPI now requires (PyPI dropped support for TLS 1.0/1.1 a while back)
• There's a corrupted pip cache entry that's causing extra issues

Here's how to fix it step by step:

1. Upgrade pip to the latest version first
   Outdated pip is the main culprit here—newer versions use TLS 1.2+ which PyPI accepts. Run this in your TensorFlow virtual environment:

   pip install --upgrade pip
   

   If you still hit SSL errors while upgrading pip, you can bypass the SSL check temporarily for this step (only do this once, as it's less secure):

   pip install --upgrade pip --trusted-host pypi.org --trusted-host files.pythonhosted.org
   

2. Clear the corrupted pip cache
   The "Cache entry deserialization failed" message means some cached data is broken. Clear the entire pip cache with:

   pip cache purge
   

3. Install opencv-python again
   Now that pip is updated and the cache is clean, try installing opencv-python once more:

   pip install opencv-python
   

   If you're still facing network issues (like slow downloads or regional blocks), you can use a trusted domestic PyPI mirror by adding -i [mirror URL] to the install command—look up reputable mirrors in your region for the exact address.

That should resolve both the SSL/TLS protocol error and the cache deserialization issue. Let me know if you run into any other snags!

内容的提问来源于stack exchange，提问作者Andy