Tkinter按钮在Python脚本正常但打包为EXE后失效求助

Hey there! I’ve run into this exact issue before—Tkinter + OpenCV apps working perfectly in Python but breaking after packaging with PyInstaller/py2exe. Let’s walk through the most likely fixes to get your Display button working again:

1. Fix OpenCV DLL Dependency Issues #

Packagers like PyInstaller often miss OpenCV’s required DLL files, which silently break camera functionality without throwing errors. Here’s how to fix it:

• Use PyInstaller’s --add-binary flag to explicitly include OpenCV’s DLLs. For example (adjust the path to match your Python installation):

  pyinstaller --add-binary "C:\Python36\Lib\site-packages\cv2\opencv_world360.dll;." your_script.py
  

• If that doesn’t work, manually copy all DLLs from your cv2 site-packages folder into the same directory as your packaged EXE.

2. Avoid Blocking the Tkinter Main Thread #

The most common culprit is running the OpenCV camera loop directly in the Tkinter main thread—this freezes the GUI event loop, making your button unresponsive. You need to offload the camera reading to a subthread, and only update the UI from the main thread using Tkinter’s after() method.

Here’s a simplified, thread-safe version of your code to test:

import tkinter as tk
import cv2
from PIL import Image, ImageTk
import threading

class CameraApp:
    def __init__(self, root):
        self.root = root
        self.camera_active = False
        self.cap = None

        # UI Elements
        self.video_label = tk.Label(root)
        self.video_label.pack(pady=10)
        self.display_btn = tk.Button(root, text="Display", command=self.toggle_camera)
        self.display_btn.pack(pady=5)

    def toggle_camera(self):
        if not self.camera_active:
            self.camera_active = True
            self.cap = cv2.VideoCapture(0)
            # Start camera reading in a daemon thread (dies when main thread exits)
            threading.Thread(target=self.update_video, daemon=True).start()
        else:
            self.camera_active = False
            if self.cap:
                self.cap.release()
            self.video_label.config(image="")

    def update_video(self):
        while self.camera_active:
            ret, frame = self.cap.read()
            if ret:
                # Convert OpenCV's BGR frame to RGB for PIL
                rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
                pil_img = Image.fromarray(rgb_frame)
                tk_img = ImageTk.PhotoImage(pil_img)
                
                # Keep a reference to avoid garbage collection
                self.video_label.tk_img = tk_img
                # Update UI in main thread with after()
                self.video_label.after(0, lambda: self.video_label.config(image=tk_img))

if __name__ == "__main__":
    root = tk.Tk()
    root.title("Camera Viewer")
    app = CameraApp(root)
    root.mainloop()

3. Add Hidden Imports to PyInstaller #

PyInstaller sometimes fails to detect all OpenCV dependencies. Create a PyInstaller spec file (run pyinstaller --name your_app your_script.py to generate it), then edit the hiddenimports line to include cv2:

hiddenimports=['cv2'],

Then repackage using the spec file:

pyinstaller your_app.spec

4. Enable Console Output for Debugging #

Packaged apps often hide error messages that would show up in the Python console. Add the --console flag when packaging to see what’s going wrong:

pyinstaller --console your_script.py

Run the EXE now—any missing modules or DLL errors will show up in the console window, which is crucial for troubleshooting.

5. Check Camera Permissions #

On Windows, packaged EXEs might be blocked from accessing the camera by system permissions. Try right-clicking the EXE and selecting "Run as administrator," or check your Windows Camera settings to ensure the app has permission to access the camera.

内容的提问来源于stack exchange，提问作者Mathilde D