A collection of I/O scripts for computer vision formats.

pip install cv-io==1.0.0


Computer Vision I/O

This python package, Computer Vision I/O which is knwon as cv_io, is a collection of existing python scripts which provides a set of I/O functions for normal and irregular image formats in computer vision.

File Format Support

File format Desciption
*.png, *.jpg, ... All normal images
*.flo Middleburry optical flow
*.dpt Middleburry floating point depth
*.pfm Freiburg floating point disparity
*.exr OpenXR hdr format
*.jxr MicroSoft JPEG-XR hdr format

Main Fucntions

Function Desciption
imread (file_name) Reads an image from "file_name"
imwrite (file_name, image) Saves the "image" array to "file_name"
imshow (image) Displays the "image" array by matplotlib
imshow (file_name) Displays an image file from "file_name"


Package requirement are listed in requirements.txt file and it will be installed through pip isntallation. If you clone the repository you may install requirements manually:

pip install -r requirements.txt

Install as a Python Package

You can also download and install the Simple Image Viewer as a python package by:

python -m pip install cv_io

or simply

pip install cv_io

Then you can simply import it in a python code and use it as follows:

import cv_io
image = cv_io.imread('samples/0512.pfm')
cv_io.imwrite('test.pfm', image)

You also have acess to the original libraries like sintel_io through the cv_io as a sub-module:

import cv_io
image = cv_io.sintel_io.depth_read ('samples/frame_0020.dpt')
cv_io.sintel_io.depth_write('test.dpt', image)

Copyright © 2020 Saeid Hosseinipoor. Released under the MIT License.