Searching photographs and images is a tricky job, is it a photo of a large chicken? Or something not quite as appropriate to show to your grandmother. Or when you search for a photo of Chris Lilley, you get one back of him with Dean Jackson, is he the tall shaven headed guy, or the one with the long flowing locks? It could be a big mistake to make when arranging that blind date.
SVG can help solve this image description problem, you can use SVG elements to chop up a photo into the relevant bits, so you can see this path of this raster is Chris and this part is Dean, you may then be disappointed but at least it'll help you avoid embarrassment when you proposition the wrong guy.
This presentation will show a couple of applications that allow you to annotate images saying which bits are what, and how SVG can then produce different ways of viewing the data, making them more accessible to the visually impaired, and far easier to search. It also enables you to filter out the parts you don't like - so now you can see the image with the large chicken, even if it's not actually a chicken, with the offending part tastefully blurred with an SVG filter.
The information is described in RDF using SVG primitives to describe the outlines of the shapes of each part you describe and wordnet and foaf vocabularies for saying who's who and what's what.
Then we'll look at an application that lets you explore a photo-album, with the searched info, so you can see a picture of the SVG Working Group, then click on Chris Lilley and see all pictures of Chris, and to see all the nose pictures just click on his nose.