ProximityMapper is a visualization utility which combines data from image files and gps output (gpx file format). By combining the timestamps of the image metadata and gps coordinate entries, the proximities of different images can be found out.
- See it in action (Flash Player 7 required)
- Download the source files (Licenced under GPL, modifying requires either .NET Framework or Mono and Flash MX 2004 Pro)
The following image shows the flow of the data from different devices.
Data is processed using a program running on .NET Framework or Mono. The program also resizes the original images and creates thumbnails for more convenient display with the Flash application, especially over the web. The following properties of the images are extracted:
- Running order of images
- Second of day
- Average amount of red, green and blue color component in the images
- Average hue, saturation and luminance in HSL color space
The visualization is done with a Flash application which can be run on any web server. The application displays image thumbnails in its window. The images can be sorted in two dimensions by any combinations of the listed properties.
The proximity of the different images can be determined by pointing any image with a mouse. Images taken near the selected image will display normally. Overexposure effect is applied to others. The strength of the effect is determined by the distance so that the further the image is taken, the brighter it is.
Images can be dragged, but other images follow, according to the proximity: closely if taken nearby.
The application works also as an image gallery. By clicking the thumbnail, display size image can be viewed. The detail view also allows quick access to next or previous image according to the sort criteria applied.
Play around with ProximityMapper to see the results. For example, sorting by hue and elevation is particularily interesting with the example images of Icelandic landscape. Another great combination is second of day and luminance. Uncheck the Proximity checkbox to see the images better.
I would like to thank RIXC and Lorna for organizing the workshop and especially Timo Arnall and Even Westvang for the gpx data used in development and some worthwhile ideas such as the HSL colors. See also their most interesting project, Time that land forgot.