9SupervisedClassificationParallelpipedClassificatioLectureNotesn

9SupervisedClassificationParallelpipedClassificatioLectureNotesn

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
1 GISc 4037 Digital Image Analysis Dr. Charles Roberts SUPERVISED CLASSIFICATION1:PARALLELPIPED CLASSIFICATION Supervised Classification was developed by frustrated image analysts who saw that unsupervised classification would never produce the information classes that they wanted for their data layers. The idea is that rather than letting the mathematics of cluster analysis determine the important spectral classes of data, we would nudge the system by telling the algorithms that certain classes were important and others weren’t. For instance, it isn’t necessary to have multiple classes of information about golf course grasses, but it might be important to map natural prairies versus planted grasses. So the image analyst would train the computer to recognize spectral classes that represent important information classes. So the classification process is supervised by a human, who trains the computer to recognize land covers by digitizing polygons that represent distinct spectral classes and important information classes. These digitized polygons are called training fields . Training Fields: the merger of Raster and Vector GISystems tools A training field is digitized using a vector tool. Remember how vector GIS works. A single pixel on a computer screen is subdivided into a grid and any location within that pixel can be referenced. So when you digitize a polygon on top of a raster computer screen, the polygon shape is smooth compared to the ragged shape of the rasters. Remember that any part of a polygon that touches a raster cell tells the computer to save the entire contents of the raster cell for statistical analysis. So a training field is really a map of a group of spatial statistics, i.e., BV’s that will be used to define a spectral signature. Consequently, if you are digitizing a
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 5

9SupervisedClassificationParallelpipedClassificatioLectureNotesn

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online