A PROJECT-BASED APPROACH TO TEACHING INTRODUCTORY IMAGE PROCESSING: ENVIRONMENTAL CHANGE DETECTION
The course uses a scaffolding approach that begins as students are introduced to image processing basics using pencils, calculators, and hard-copy pixel arrays. Image processing software (MultiSpec) is introduced in tutorial exercises that lead to unsupervised classification of Landsat imagery followed by change detection. Students use commercial image processing software in a tutorial developed at UWG to perform multi-temporal land cover classifications and environmental change detection, followed by raster to vector conversion and importation into a GIS. Students repeat the process they have successfully applied in the tutorial over an area of interest they select, and prepare a final presentation highlighting environmental change over the chosen time frame and location.
Several semesters of implementation have demonstrated that the scaffolding approach builds on repeated student successes, giving students the confidence to generate and complete their own projects. Students' selection of sites with which they are familiar generates diverse questions and insights, placing image processing within a familiar and interesting context in which students learn about both environmental change and image processing techniques, while generating and answering questions of importance to them and sharing what they learn in a professional format. This presentation will illustrate exercises and products from all phases of the class, including resource sites that 6 16 level educators can use to develop similar environmental change exercises in diverse disciplines including Earth and Environmental Science. The presentation will also display examples of final student project results and presentations.