DICTA

The 2007 edition of Digital Image Computing Technologies and Applications (DICTA) was held near Adelaide, South Australia, in the seaside community of Glenelg. DICTA is hosted by the Australian Pattern Recognition Society (APRS) and has been held approximately every other year in various cities in Australia and New Zealand. DICTA 2007 was the ninth DICTA conference.

DICTA 2007 comprised approximately 80 contributed papers and six keynote addresses and a dinner speaker. Just over a hundred delegates attended, approximately one fourth from outside Australia and New Zealand.

The keynote addresses reflected the major themes of the conference. Edwin Hancock of the University of York spoke on graph spectral methods for learning shape categories, and Maria Petrou of Imperial College continued on the theme of learning by presenting her work on the tower of knowledge for learning in computer vision. Core image and video processing was represented by keynote speaker Jan Flusser of the Institute of Information Theory and Automation, Czech Republic, who provided an overview talk on image fusion, and by Ian Reid of Oxford, who described methods for creating new views of a scene from video streams and augmenting video with computer graphics using real-time camera tracking.

Numerous applications of image analysis were presented ranging from well established areas such as face recognition and finger print analysis to more novel directions such as identification of swimming pools, classification of birds from shape, and the counting of sea stars. The single largest application field was to biology and medicine. Work here was represented by keynote speaker Michael Brady of Oxford who spoke on image analysis related to cancer and by Geoffrey McLennan of the University of Iowa who considered the vast integration of image analysis in the future of medicine under the banner of a new field called Eidomics. Both speakers emphasized the need for researchers in medicine and image analysis to work as a team.

DICTA 2007 was preceded by a workshop on vision in human computer interaction (Vis-HCI) held at the same venue on December 2. A session of contributed papers on Vis-HCI was included in DICTA to reinforce interaction between researchers in this area and general image analysis.

Due to a formal request and promised funding support from one of DICTA’s major sponsors, it was decided at the APRS general meeting that DICTA would become an annual rather than biennial conference. The next conference in the series is thus expected to be held in Canberra in 2008.