通过Numpy RGB阵列创建Vips图像以获得大图像

时间:2018-11-28 05:10:07

标签: python numpy vips

我试图用PIL / Pillow输出生成的大图像,但是当图像尺寸变大时会中断。

因此,根据我在SO上阅读的内容,我正在尝试使用Vips。

我生成的数据是numpy的RGB值数组。我想将其转换为Vips图像,以便保存。但是我不知道如何将像素数据转换为Vips。

import numpy
import gi
gi.require_version('Vips', '8.0')
from gi.repository import Vips

WIDTH=32768
HEIGHT=32768
UCHAR=Vips.BandFormat.UCHAR

# Create an RGB black image
black_space = numpy.zeros( ( WIDTH, HEIGHT, 3 ), dtype=numpy.uint8 )

# this doesn't work
vips_image = Vips.Image.new_from_memory( black_space, WIDTH, HEIGHT, bands=3, format=UCHAR )
vips_image.write_to_file( "space_32k.tiff" )

当然,创建Vips图像时它会失败并显示错误:

Traceback (most recent call last):
  File "./bad_vips.py", line 14, in <module>
    vips_image = Vips.Image.new_from_memory( black_space, WIDTH, HEIGHT, bands=3, format=UCHAR )
TypeError: Item 0: expected int argument

是否有一种转换numpy数组的方法,使其可与Vips一起使用?

我也尝试传递black_space.data,但随后得到:

NotImplementedError: Item 0: multi-dimensional sub-views are not implemented

1 个答案:

答案 0 :(得分:1)

您正在使用旧的libvips Python接口-现在有一个更新的接口,它好很多了:

https://github.com/libvips/pyvips

此处的文档:

https://libvips.github.io/pyvips/

a section关于链接libvips和numpy。您的示例为:

import numpy
import pyvips

WIDTH = 100
HEIGHT = 100

# Create an RGB black image
black_space = numpy.zeros((WIDTH, HEIGHT, 3), dtype=numpy.uint8)

# reshape into a huge linear array
linear = black_space.reshape(WIDTH * HEIGHT * 3)

vips_image = pyvips.Image.new_from_memory(linear.data, \
        WIDTH, HEIGHT, bands=3, format="uchar")

vips_image.write_to_file("huge.tif")

reshape是免费的(我认为),因此应该高效。回购中还有一个示例程序:

https://github.com/libvips/pyvips/blob/master/examples/pil-numpy-pyvips.py