This article was produced and financed by Oslo Metropolitan University
Let the sun organise your photos
Fed up with the chaos in your digital photo collection? New research shows that the sun and the clock can sort out your pictures.
Denne artikkelen er over ti år gammel og kan inneholde utdatert informasjon.
Oslo Metropolitan University
We now take more pictures than ever before. Many own a digital camera. Our computers are full of photos taken on holidays, at Christmas, at birthday parties and weddings.
Many photographers have neither the time nor the patience to tag their photos manually with text labels. The result is that the pictures often end up stored in folders named ‘Holiday 2011’ or 'Christmas 2010’.
“It can be fun taking photos, but not as much fun having to organise them afterwards,” says Professor Frode Eika Sandnes at the Oslo and Akershus University College of Applied Sciences (HiOA). He has developed a new method that automatically detects where you have been by analysing the photos in your digital camera.
Light and time
Eika Sandnes’ method, which was recently published in an academic journal, uses the level of light and the time of day when a photo was taken to organise your images. There is often more light around midday than at late afternoon or early morning. Based on the level of light in a collection of images, it is possible to estimate approximately which ones were taken around midday and which ones were taken around sundown.
The number of hours of sunlight depends on our geographical location and the time of year. In the northern hemisphere the days are shorter during the winter and longer during the summer:
“Basic astronomical calculations can be used to determine a geographical location when we know approximately when the sun is at its highest and when it sets,” says Eika Sandnes.
Fast and automatic
The image sorting process would be fast and automatic, and would provide you with a systemised overview:
“Among other things, you would obtain a better overview of large collections of images, and the method simplifies searches for specific images," says the Professor.
This can also help a blind computer user because the system will automatically generate image descriptions.
GPS is cumbersome
A number of studies have been conducted to find out how images can best be organised, but so far there are few software programmes around that can automatically sort them.
“There are some programmes that can recognise landmarks and connect images to a specific location, but they don’t work on locations that have no such landmarks,” says Eika Sandnes.
GPS can also be used to tie images to locations, but then you need to have GPS in your camera, and this is not currently the case. Also, GPS is energy-intensive, expensive, and takes a long time to connect to, all of which is impractical when you want to take a spontaneous picture.
Simple to use
Frode Eika Sandnes’ method is simple, and works as long as the camera’s clock is functioning.
“Most people don't bother about the clock in their camera, and don't do anything with it. Also, it runs on its own battery," he says. He envisions a method that can be used in new software and that can be integrated with today’s image recognition technology for sorting images by events.
Translated by: Sonja Balci
- Frode Eika Sandnes, Where was that photo taken? Deriving geographical information from image collections based on temporal exposure attributes, Multimedia Systems, Vol. 16, No. 4-5, pp. 309-318, 2010. (Abstract)