We would like to state that we have no commercial or scientific interests. All we want is to help the community to improve research and in this way help conservation of endangered species. We do this (in our spare time) simply because we enjoy it and want to contribute.
All work is done for free, but as owners of small enterprises, working on the tool also means no income whatsoever. Donations or grants for future development are most welcome.
Follow us on
Introduction to I3S Spot
I3S Spot (previously named Manta) is the second application of the I3S family and is basically an extension of I3S Classic. In Classic annotation is limited to the location of the centre of the markings, in Spot you also annotate the shape of the marking by drawing the closest fitting ellipse around it. In this way information about shape and size is used, making it possible to exclude potential matches in an early stage using simple constraints on shape. These constraints are the ratio between the long and short radius of the ellipse, and size (the area of the ellipse).
The matching algorithm is identical to I3S Classic except that I3S Spot will increase the score as spot pairs differ in size or shape. Lower scores mean better matches. To be able to calculate an accurate score, a sufficient number of spots is required, typically 12 – 30. If more there are more than 30 spots available for annotation use the biggest or most prominent first. Annotating more than about 30 spots may cause loss of accuracy, but will certainly increase search time.
For many species with a stable identification area (i.e. not very susceptible to curving) and sufficient spots, e.g. newts, experiments on a large database (2627 photos from 1075 individuals) show that when using 2 reference images, in 96.5% of the cases, the corresponding individual is in the top 10 of best matches shown by I3S Spot.
However, if the number of spots is either low, or the animals are prone to curvature in the identification area (e.g. manta rays) performance may be less accurate.
I3S Spot v4.0 allows you to set some internal parameters. These parameters were hard coded in I3S Manta v2.1. For explanation of these parameters you are referred to the manual coming with the installation. You can also contact us directly via e-mail.
New features in version 4.0 are:
- Improved user interface showing the results. Now it is possible to view the image from the result list. With AutoZoom you directly zoom in on the area of interest. Finally, you can tell I3S to show either file names or directories in the result list.
- Significant improvements in accuracy.
- Speed of the recognition improves by a factor 3 because I3S now uses the parallel processing power of your computer.
- The possibility to use and manage multiple databases.
- Easy and user-friendly definition of your own meta data structures.
- Various tools to evaluate the performance of I3S on your database and to compare two databases.
- I3S comes with the possibility of automatic updates.
- Key parameters of the recognition algorithm can be tailored for your specific data set.
- Various minor bug fixes and improvements.
I3S comes with a simple conversion tool for your current database.