A Python package for fast and robust Image Stitching.

Based on opencv's stitching module and inspired by the stitching_detailed.py python command line tool.

use the docker image

or pip to install stitching from PyPI.

pip install stitching

The command line interface (cli) is available after installation

stitch -h show the help

stitch * stitches all files in the current directory

stitch img_dir/IMG*.jpg stitches all files in the img_dir directory starting with "IMG" and ending with ".jpg"

stitch img1.jpg img2.jpg img3.jpg stitches the 3 explicit files of the current directory

Enable verbose mode with stitch * -v. This will create a folder where all intermediate results are stored so that you can find out where there are problems with your images, if any

If you are familiar with Docker and don't feel like setting up Python and an environment, you can also use the openstitching/stitch Docker image

docker container run --rm -v /path/to/data:/data openstitching/stitch:{version} -h

You can use the CLI as described above (read "current directory" as "/data directory")

You can also use the Stitcher class in your script

from stitching import Stitcher
stitcher = Stitcher()

Specify your custom settings as

stitcher = Stitcher(detector="sift", confidence_threshold=0.2)

or

settings = {"detector": "sift", "confidence_threshold": 0.2}
stitcher = Stitcher(**settings)

Create a Panorama from your Images:

• from a list of filenames

panorama = stitcher.stitch(["img1.jpg", "img2.jpg", "img3.jpg"])

• from a single item list with a wildcard

panorama = stitcher.stitch(["img?.jpg"])

• from a list of already loaded images

panorama = stitcher.stitch([cv.imread("img1.jpg"), cv.imread("img2.jpg")])

The equivalent of the --affine cli parameter within the script is

from stitching import AffineStitcher
stitcher = AffineStitcher()
panorama = stitcher.stitch(...)

The equivalent of the -v/--verbose cli parameter within the script is

panorama = stitcher.stitch_verbose(...)

For questions please use our discussions. Please do not use our issue section for questions.

Read through how to contribute for information on topics like finding and fixing bugs and improving / maintaining this package.

This package provides utility functions to deeply analyse what's happening behind the stitching. A tutorial was created as Jupyter Notebook. The preview is here.

You can e.g. visualize the RANSAC matches between the images or the seam lines where the images are blended:

This package was developed and used for our paper Automatic stitching of fragmented construction plans of hydraulic structures

Apache License 2.0