Conole, Evans and Simms, the authors of Chapter 13, Use and Reuse of Digital Images in Teaching and Learning, have provided a coherent and well structured description of the issues involved in use of digital imagery in education. The authors have done a good job of describing the aspects of granularity and context and how they relate to utilization of images in education.
The authors of Chapter 13, Use and Reuse of Digital Images in Teaching and Learning, begin with a very useful image mapping exercise, which describes a selection of image types. Then they group these examples into a table of image categories. I would add two additional image classes to the table, which are very important in building image libraries in support of learning. From the Image mapping exercise they are: the 3D Model and the virtual reality model. These two examples are important because they typify the computer image primitive markup, which generates an image (as in a 3D modeling system or a computer graphic metafile) and the rendered or generated virtual reality image. As the semantic web evolves, with its goals of machine and human readability and machine understanding of web documents using XML, these two synthetic image types will take on ever more importance.
Some of the biggest challenges in image reuse have to do with classification, cataloging, and discovery. In the chapter, two important JISC image capture and cataloging projects are mentioned. I would like to add mention of the ongoing work at Columbia University on WebSEEK, and VisualSEEK, Berkeley Blobworld, Stanford VISION, and Penn State's SIMPLIcity.
Of these projects, the Content Based Image Retrieval (CBIR) projects are interesting for educators because they seek to provide richer and deeper semantic classification schemes. Image classification is usually done in one of three ways:
Objective cataloging - indexes and retrieves images by primitive features such as color, texture, shape or the spatial location of image elements. It uses features, which are both objective, and directly derivable from the images themselves, without the need to refer to any external knowledge base.
Logical cataloging - indexes and retrieves images by derived features, involving logical inference about the objects depicted in the image.
Abstract cataloging ? indexes and retrieves images by abstract qualities, involving high-level reasoning about the meaning and purpose of the objects or scenes depicted.
In addition to the hundreds of thousands of images cataloged in the research systems, both commercial and non-commercial search engines now provide search strategies and access to hundreds of thousands of additional images. Most of these search engines use objective keyword indexes as their main form of retrieval. The following specialized cataloging schemes for images are well known: The Getty Archive, the RPI Art and Architecture Thesaurus (AAT), the Vienna classification for trademark images [World Intellectual Property Organization, 1998], and the Library of Congress Photo Classification Scheme.
In the section on images and pedagogy describing the importance of images in education, references to Howard Gardner's Multiple Intelligences, recent cognitive and neuroscience research, findings in left-brain right-brain studies, gender differences in perception, learning styles and visual literacy could add an important dimension. From the I Ching to 20th century Semiotics, the role played by images in culture and education has been central. As part of art studies, visual literacy, visual grammars, and visual syntax are important subjects for future study.
A short review of image technologies and formats would be useful in understanding the implications of proprietary and open image formats for use and reuse of images. It is also important to point out that there are two new image formats, which have particular significance because they contain features, which are optimized for the Internet and put them in a class by themselves.
The Portable Network Graphics (PNG) format was designed to replace the GIF with a royalty-free format. PNG controls its own brightness using gamma correction. In a cross-platform environment GIFs and JPEGs don't always look the same on every platform. With PNG, you don't have to worry about that because it is self-correcting. Another important thing about PNG is that it uses alpha channels for improved anti-aliasing and control of transparent areas within an image. Like GIF, PNG uses lossless compression.
W3C Scalable Vector Graphics (SVG) is a new graphical presentation language based on XML. SVG is a language for describing two-dimensional graphics in XML. SVG drawings can be dynamic and interactive. By leveraging SVG together with other web technologies, powerful interfaces can be built that work consistently across all sorts of browsers and present high quality, scalable imagery.
Other visual coding and XML markup schemes like MathML will replace static images currently used in learning systems with dynamic markup and data driven imagery.
The authors have done a good job of describing the aspects of granularity and context and how they relate to utilization of images in education. There is one other distinction, which should be made when considering the use and reuse of digital images. Hypertext creates an extended dimension for linking to, from, and into text documents, which transcends their reusability as standalone documents. The SVG image format and other commonly used hyperlink image maps transcend the reusability of individual images. When regions of an image are clickable, the image forever transcends being objective, passive or neutral. Its structure is active and its parts have meaning.. Full implementation of W3C XLINK, XPATH, and XPOINTER capabilities promise to create altogether new ways of perceiving, understanding, manipulating and using images on the Web.
In the section on Intellectual Property (IP) rights and images it would be worth mentioning the emerging technologies available for watermarking and embedding rights information into existing image formats.