如何设置cv2.addWeighted的alpha和beta？循环权重值失效问题求助

Hey there! Let's break down why your image blending loop isn't working correctly and fix it step by step.

Common Problems & Solutions #

1. Integer Division (Unintended Non-Float Values) #

If you're running this code in Python 2, dividing two integers like i/30 will perform integer division—meaning 1/30 becomes 0 instead of the floating-point 0.033 you need for proper blending. Even in Python 3, explicitly casting to float avoids edge cases and makes your intent clearer.

Fix: Replace i/30 with float(i)/30 or i/30.0 to ensure alpha and beta are floating-point values.

2. Mismatched Image Dimensions/Channels #

cv2.addWeighted requires both input images to have the exact same height, width, and number of channels (e.g., both 3-channel RGB images). If your images differ in size, the function will either throw an error or produce broken results.

Fix: Resize one image to match the other's dimensions before blending:

# Resize src2 to match src1's width and height
src2 = cv2.resize(src2, (src1.shape[1], src1.shape[0]))

3. Missing Output Folder #

If the newfolder directory doesn't exist, cv2.imwrite will fail silently (no error message, but no images saved). You need to create the folder first.

Fix: Use Python's os module to check and create the folder:

import os
os.makedirs('newfolder', exist_ok=True)

Full Fixed Code #

Here's the complete code incorporating all fixes:

import cv2
import os

# Create output folder if it doesn't exist
os.makedirs('newfolder', exist_ok=True)

# Read two images
src1 = cv2.imread('image1.png', cv2.IMREAD_COLOR)
src2 = cv2.imread('image2.png', cv2.IMREAD_COLOR)

# Handle cases where images fail to load
if src1 is None or src2 is None:
    raise ValueError("One or both images failed to load. Check file paths!")

# Ensure both images have the same dimensions
if src1.shape != src2.shape:
    src2 = cv2.resize(src2, (src1.shape[1], src1.shape[0]))

# Blend images across different weights
for i in range(1, 30):
    alpha = float(i)/30
    beta = 1.0 - alpha
    dst = cv2.addWeighted(src1, alpha, src2, beta, 0.0)
    # Save the blended image (using f-string for cleaner syntax)
    cv2.imwrite(f'newfolder/image{i}.png', dst)

Quick Extra Tip #

I used an f-string (f'newfolder/image{i}.png') for cleaner file naming—this works in Python 3.6+. If you're on an older Python version, stick with your original syntax: 'newfolder\\image'+str(i)+'.png'.

内容的提问来源于stack exchange，提问作者Fede