10.6. Assignment 2 – Comparison of classification algorithms

Use the knowledge/experience gained from practical’s under section 10.3, 10.4 and 10.5 to help you process the data and produce the report for Assignment 2.   N/B: Information on how to perform accuracy assessment can be found in Section 11.1 and 11.2. 

Assignment 2:

Task: Compare and contrast the performance of three distinct classification approaches of your choice (i.e. one unsupervised, one supervised and one object based classification method). The report should include, but not limited to: introduction/literature review, methods, results, discussion and conclusion.

Data source: Download a cloud free Landsat 4/5 image over Southampton, UK urban area captured on the 3rd June 2006 from Earth Explorer(https://earthexplorer.usgs.gov/) . Example of how to download data from Earth Explorer, layer stack the image and subset the image can be found here.

Before performing classification on the image,  pre-process the image as appropriate (e.g. subset to region of interest).

The total word count is 2000(excluding references).

Some references

Congalton, R.G., (1991). A review of assessing the accuracy of classifications of remotely sensed data, Remote Sensing of Environment, 37(1): 35-46

Foody G.M. (2002) Status of land cover classification accuracy assessment. Remote Sensing of Environment, 80, 185-201

Foody G.M. (2004). Thematic map comparison: Evaluating the statistical significance of differences in classification accuracy. Photogrammetric Engineering and Remote Sensing, 70, 627-633.