ABS project was created with the intention of detecting blurred and/or flared images.
This project const of 5 files, 2 for blurred images detection and 2 for flared images detection:
-
blurry-or-no.py and flared-or-not.py: Takes an image and analyse it to determine if it's blurred or flared respectively.
-
blurrytest.py and flaredtest.py: This two files was created with the intention of analyse and verify these python programs.
Each of this programs takes all images in a given directory and analyse them using a window as a interface to show the image and result of the analysis. It's more convenient for analysing a bunch of images without invoking the command for every single image.
"validation.sh": It's a Linux bash script made to validate the python scripts.
Getting started
To run and test this commands you can use the python console or run them directly in your Ubuntu command line interface (if you already have installed python in your linux machine).
Requisites
This python scripts are made with the following configuration system:
- Ubuntu 16.04.5 LTS
- Python 2.7.12
- OpenCV 2.4.9.1
Once you get your installation done (check this howto if you ineed to install it https://docs.opencv.org/3.4.1/d2/de6/tutorial_py_setup_in_ubuntu.html ) you are ready to run this programs. Here are some examples.
-
To run flared-or-not.py just type: ./flared-or-not.py -i /path_to_your_data/image.jpg -t 115
-
To run flared-or-not.py just type: ./flared-or-not.py -i /path_to_your_data/image.jpg -t 73
-
To run blurrytest.py just type: ./blurrytest.py -i /path_to_your_data//directory_containing_images -t 115
-
To run flaredtest.py just type: ./flaredtest.py -i /path_to_your_data/directory_containing_images -t 73
Validation script use
Tu verify and validate the python algorithms with any set of images follow this set of instructions:
- Download the files
- Create or copy the directory "training-data" in /home//
- Execute the script (./validation.sh)