调用cv2.addWeighted时报'Tuple'对象不可调用错误的求助

First, let's clarify your situation: You wrote this OpenCV code based on a tutorial, and when you run it in Spyder 3.2.8 (Python 3.6.4, Windows 10), you hit a TypeError: 'tuple' object is not callable error. But when you run the exact same code in the Windows Command Prompt, it works perfectly.

Here's your code, with a quick fix to the import syntax (you had two imports on one line, which is invalid Python—probably a paste error):

import cv2
import numpy as np

img1 = cv2.imread('3D-Matplotlib.png')
img2 = cv2.imread('mainsvmimage.png')

# add = img1 + img2
# add = cv2.add(img1, img2)
# added all pixel values together
# (155, 211, 79) + 50, 170, 200 = 205, 381, 279
# which becomes (205, 255, 255) since 255 is the max
# add = cv2.add(img1//2, img2//2)

print(img1.shape)
print(img2.shape)

weighted = cv2.addWeighted(
    src1=np.array(img1), 
    alpha=0.6, 
    src2=np.array(img2), 
    beta=0.4, 
    gamma=0
)

cv2.imshow('image', weighted)
cv2.waitKey(0)
cv2.destroyAllWindows()

Why This Happens Only in Spyder #

Spyder uses an interactive IPython kernel that keeps track of all variables and function definitions from every code run you do in the same session. The error you're seeing means at some point, you accidentally replaced the cv2.addWeighted function with a tuple without realizing it.

For example, maybe you ran a line like addWeighted = (100, 200) earlier in the session, or you reused a variable name that conflicted with the OpenCV function. Once that happens, Spyder remembers that "addWeighted" is a tuple, not a function, until you reset the environment. The Command Prompt doesn't have this issue because it runs code in a fresh, clean environment every time you execute a script.

How to Fix It #

1. Restart the Spyder Kernel (Quickest Fix) #

This is the easiest way to wipe the slate clean:

• In Spyder, go to the Consoles menu and select Restart kernel (or press Ctrl + . as a shortcut).
• Once the kernel restarts, run your code again from the top. This will restore all the original OpenCV function definitions and clear any conflicting variables.

2. Clear the Namespace Manually #

If you don't want to restart the kernel, run this IPython magic command first to delete all existing variables:

%reset -f

Then re-import cv2 and numpy, and run your code. This ensures no leftover variables are messing with your function calls.

3. Check for Accidental Overwrites #

Take a quick look at any previous code you ran in the same Spyder session. Make sure you didn't use variable names that match OpenCV functions (like addWeighted—use something like weighted_image instead). Even a single line where you assign a tuple to that name can cause this error.

4. Fix the Import Syntax #

Don't forget to split your import lines into two separate lines (as shown in the corrected code above). While you said the Command Prompt ran it, that's likely a typo in your original paste—fixing this prevents any unexpected parsing issues in Spyder.

Final Tip #

Spyder's interactive setup is great for testing code incrementally, but it can lead to these kinds of namespace conflicts. If you ever run into an error that doesn't make sense (especially if it works in a fresh environment like the Command Prompt), restarting the kernel is usually the first thing to try.

内容的提问来源于stack exchange，提问作者Q.H.