This package provides functionalities to ease images handling for computer vision projects, written in python.
In this repository you will find the following python scripts :
-
📁 display_tool/ : This tool aims to help visualizing the outputs of some algorithms along with the corresponding inputs. It is especially usefull for benchmarks. More information is available in display_tool/readme.md.
-
📁 images_comparison/ : In this repository you will find several scripts to measure the proximity and similarity between two images. This is a type Image Quality Assessment algorithms for which we always have the ground truth, in this case we say that we are dealing with full-reference iqa.
- 📝 crop_central.py : This python script enables to central crop all images in a given input folder if their shapes exceed a certain width and height. The cropped images are saved into the given output directory.
python central_crop.py --folder_path *path/to/images* --output_dir *path/to/folder/to/save/images* --width_max *maximum_width* --height_max *maximum_height*
- 📝 crop_select.py : This file will allow you to crop an image based on the selection you will input when script launched.
python crop_select.py --image_path *path/to/image*
- 📝 dataset_cleaning.py : this script will look at all images in the given folder_path and rename all images with the following standard foldername_000XX.png. The script also outputs a csv file that contains all images former names and new ones along with the images sizes.
python dataset_cleaning.py --folder_path *path/to/folder*
- 📝 histogram_equalization.py : to enhance the contrast of an image, equalizing its histogram might be a good alternative. This script will output an histogram equalized image corresponding to the given input
python histogram_equalization.py --image_path *path/to/image*
All required python packages to run any script from this repository are listed in requirements.txt. We advice using Python3.
blablabla