‘We should no longer see imagery as something separate from GIS,’ says Lawrie Jordan. ‘Rather, we should see GIS and imagery as complementing and completing each other.’
Jordan speaks with some authority. He recently joined software vendor ESRI as a director of imagery enterprise solutions, also serving as special assistant to ESRI founder Jack Dangermond. But Jordan says he sees his role with the company as more of an imagery evangelist.
He has come a long way since he and Dangermond attended the same college back in the late 1960s. He founded Earth Resource Data Aquisition Systems in the late 1970s. The company shot to prominence on the basis of a revolutionary new software product that was capable of image manipulation on an 8-bit desktop computer with a Z80 processor.
In 1991, this morphed into ERDAS Imagine, which ran on an IBM PC.
He and Dangermond have shared a close professional relationship and for many years, Imagine was the de facto image handler for the entire ESRI product suite.
That came to an end when Jordan sold ERDAS to Leica Geosystems Geospatial Imaging. Leica later resuscitated the name and used it for its geospatial imaging division.
So what keeps him passionate? Geospatial imaging, he says, is something that has the power to make a difference.
He says the modern imaging world is founded on Moore’s Law, the doctrine credited to Intel founder Gordon Moore, which plots the growth in computing power against time. Strictly speaking, Moore was talking about the number of transistors that can fit on a silicon wafer. Jordan’s point is that the same techniques allow us to increase the storage capacity of disk drives, the amount of RAM on board, and the speed with which imagery can be processed.
Modern imaging is also founded on compression technology, which allows faster processing of bigger images. Jordan is a big fan, of course, but he is cautious about lossy compression techniques and non-standards-based approaches to the problem. The outcome is a world where it is possible to integrate imagery and vectors in a seamless manner. Complex tasks such as image processing and automated feature extraction can now be streamlined into simplified workflows and templates. This allows us to encapsulate the domain-specific knowledge of experts and serve meaningful results to an enterprisewide constituency. ‘It’s the illusion of simplicity,’ he says. A major benefit is that GIS will increasingly be optimised for image handling – and as a consequence, the historical separation between imagery and GIS analysts will disappear. The latest incarnation of ESRI ArcGIS suite is a case in point. It now employs an image extension to ArcGIS Server. Images are stored in their native format on the server and then processed in real time as the image, or portions of it, are required. The processing itself is undertaken by a ‘service’, which only needs to be defined once. It is a technique that makes the metadata searching, discovery, and serving of images dramatically faster than older methods, he says. This modern Services Oriented Architecture makes it possible to engage in new and more useful ways of cataloguing images. It even makes it possible to organise them into federations of catalogues.Why does all this matter? Jordan says the greatest problem of our time is sustainability – ensuring that we have an Earth, a way of life and a civilisation to pass on to our children. There is an increasing realisation that national, economic, and environmental security are fundamentally intertwined. Geospatial information is a bridge. It helps us understand these dynamics. It will be an integral part of our stewardship of the Earth.
There are still enormous challenges to be overcome. The sheer volume of imagery that might be required and our ability to analyse it is one. Temporal studies that allow us to visualise the way in which patterns are changing, remain difficult. The restrictions on image gathering imposed by governments is a real problem. We still don’t understand how to do threedimensional representations.
Of Google, Jordan describes the advent of computer globes as ‘a gift to GIS’. ‘They raise our collective level of awareness; they set the stage for new and exciting ways for presenting all types of geospatial knowledge,’ he says.
Jon Fairall interviewed Lawrie Jordan in Hyderabad on 12 February 2009.
