Image-A-Day

  • 2009
  • ongoing
  • Website: Image-A-Day
  • This software is free software, licensed under the GPL 3.

Every day in 2009 I added code to a software program that "collages" information from my day, including photos, text, website browsing history, music I listened to, number of key presses and mouse clicks, etc. It transforms this information into a single image that is then posted onto a website. Other participants also post their image every day to the website, the only rule being that each participant must upload an image for each day.

Image-A-Day example images

Description

Image-A-Day monthly view

This work came out of desire to challenge myself to create art every day for a year. I wanted to write software but I also wanted a project that my friends and family could join. Thus was born the Image-A-Day Journal: a website where participants could upload and share an image every day. The only rule was that you had to upload an image each day.

My software became a "collager": I "feed" it images and text from my day and it manipulates and positions them. I eventually added a number of features that collaged additional details about my day: websites I visited, number of key presses and mouse clicks, music I listened to, news and weather, etc. Through the medium of digital collage this information is shared with the other members of the project.

The Journal explores how people change as their lives become more documented and as they participate and share in artistic practice every day.

Features

Currently implemented features:

  • IMAGES

  • Process images (jpg, png)

    Thanks to Python Imaging Library

  • Detect common colours in images and set border, text, and background gradient from these
  • Smart image scaling to adapt to different sized images
  • Smart image scaling to adapt to number of images to collage and their relative sizes
  • Place "recurring" images behind today's unique images

  • TEXT

  • Process text files
  • Process html files

    Thanks to Beautiful Soup

  • Find common words in text

    Thanks to the Natural Language Toolkit (nltk)

  • Find titles in text

  • TRACKING

  • Track websites visited and extract common words (from Firefox)
  • Track songs listened to and display names and images of albums (from Last.FM)

    Thanks to python-lastfm

  • Track tasks completed (from RememberTheMilk.com)

    Thanks to Sridhar Ratnakumar and his pyrtm library

  • Track time spent on computer

    Thanks to Project Hamster

  • Track number of mouse clicks and key presses

    Thanks to Workrave

  • News headlines [for Toronto and copyright issues] (from Google News)

    Thanks to the Universal Feed Parser

  • Current weather (Yahoo weather or Google weather)

    Thanks to the Python Weather API

  • Take a snapshot from webcam each time the software is run

    Thanks to Fredrik Portström and the python-v3l2-capture project

  • Display time that final run of software started (to encourage me to go to bed earlier)

  • MISCELLANEOUS

  • Display code changes in spiral path around border (from baazar)
  • Display how much time taken working on today's code
  • Display how much time taken executing/processing today's collage
  • Plot a line graph of tasks completed, mouse clicks and key presses

features diagram

Development

The collager software is written in Python 2.6. It relies on a large number of free and open source libraries and other projects, including but not limited to, and in no particular order:

The website is a mix of PHP and javascript.

Thanks to all the people who shared their code and made this project possible.