Stitching

Stitch images to form a montage.

Requirements

Java development kit

Python 2.7.12

pip

numpy

$ pip install numpy

scipy

$ pip install scipy

click

$ pip install click

skimage

$ pip install skimage

python-bioformats

$ pip install python-bioformats

matplotlib

$ pip install matplotlib

Additional in Windows OS: Visual C++ 9.0

Installation

$ git clone https://github.com/CellProfiler/stitching.git

$ cd /path/to/stitching

$ pip install -e .

Info

Users may notice that when opening and zooming the images in IDEAS (and also in ImageJ, IrfanView) the images appear to be higher resolution, but this is because the software has an adjustable DPI setting, which uses interpolation to scale the images up (adding pixels) for display and publication.

Use

$ python stitching -o path/to/OUTPUT_DIRECTORY path/to/IMAGE

Generates per-channel tiled images from IMAGE saved to OUTPUT_DIRECTORY. Each image in IMAGE has shape 55px by 55px and is padded with random noise. Files are named "ch1.tif", "ch2.tif", ..., one for each channel.

Optional:

--image-size

Set the window size of the tile images (default: --image-size 55). If user sets a size bigger than the original images, each of the tile images will be padded with its own background. If user sets a size smaller than the original images, each of the tile images will be cropped toward its center.

--montage-size

Set the size of the final montage (default: --montage-size 30). If the total number of original images could not fill up the desired montage size, the montage will be padded with black background. For example: Number of total images: 1000 User setting: --image-size 20 --montage-size 30 The result would be 2 montages at the size of 600x600; one montage is filled up with 900 (30x30) tile images, one montage contains 100 tile images and remaining space is filled up by black background.