如何使用魔杖扭曲图像

我正在尝试做同样的事情this https://stackoverflow.com/questions/52090350/how-to-insert-image-in-a-mock-up老问题但在python using wand.

到目前为止我已经做了以下事情：

from wand.image import Image
from wand.display import display

# Paths to your input images
background_path = 'background.png'
source_path = 'source.jpg'
output_path = 'output.jpg'

# Open the background image
with Image(filename=background_path) as background:
    # Open the source image (source)
    with Image(filename=source_path) as source:
        # Define the distortion points
        distortion_points = (
            0, 0, 0, 0,
            265, 0, 540, 85,
            265, 333, 594, 426,
            0, 333, 342, 446
        )

        # Apply the perspective distortion to the blank image
        source.distort('perspective', distortion_points)

        # Composite the background image onto the distorted source
        background.composite(source, 317, 99)

        # Save the resulting image
        background.save(filename=output_path)

display(Image(filename='output.jpg'))

The flower图像扭曲且像素未对齐。

知道如何仅在 python 中实现相同的效果吗？

EDIT:

这是目前的状态

current result

expected result

background image

源图像

Thanks

我认为你需要这样做更多：

#!/usr/bin/env python3

from wand.image import Image
from wand.display import display

# Paths to your input images
background_path = 'background.jpg'
source_path = 'source.jpg'
output_path = 'output.jpg'

# Open the background image
with Image(filename=background_path) as background:
    # Open the source image (source)
    with Image(filename=source_path) as source:
        # Define the distortion points
        distortion_points = (
            0, 0, 317, 99,
            265, 0, 540, 85,
            265, 333, 594, 426,
            0, 333, 342, 446
        )
        # See https://docs.wand-py.org/en/0.6.7/guide/distortion.html
        source.virtual_pixel = 'transparent'
        source.artifacts['distort:viewport'] = f'{background.width}x{background.height}+0+0'

        # Apply the perspective distortion to the blank image
        source.distort('perspective', distortion_points)

        # Composite the distorted source onto the background
        background.composite(source, 0, 0)

        # Save the resulting image
        background.save(filename=output_path)

display(Image(filename='output.jpg'))