9MinDistPowerpnts

9MinDistPowerpnts - – Many clusters, small cluster...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon
Review of the Supervised Classification Process • Appropriate classificaiton scheme adopted • Training fields that represent all classes of spectral information selected – Nearly homogeneous in tone – Typical, not unusual examples of land-cover – Number of pixels in training field should be 10 x the number of bands – Consideration of temporal and spatial considerations
Background image of page 1

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

View Full DocumentRight Arrow Icon
Review, Continued • Statistics from training fields extracted and analyzed • Appropriate classification method chosen • Image classified • Accuracy Assessment – Spatial sample – Ground truthing – Accuracy asessment – Analysis of Structure of errors – Reclassify: Second iteration
Background image of page 2
Ideal sequence for Training Field Selection • Assemble ancillary data – DRG, DOQQ, county hwy maps – Examine image, find landmark features • Unusual spectral signatures • Typical spectral signatures • Unsupervised to determine broad spectral areas
Background image of page 3

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

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

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

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

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

View Full DocumentRight Arrow Icon
Background image of page 8
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: – Many clusters, small cluster radiuses, etc. – Take and evaluate training fields – Ground truth training field sites Minimum Distance to Means Classification • Advantages – Slower than parallelpiped, faster than maximum likelyhood – Good to use when training fields are small or not very well defined – Entire image should be classified with no holes How minimum distance works • Mean vector from training fields • Each pixel in image compared to training field mean vectors • Minimum distance calculated by one of two methods Euclidean Distance between Point a and training field b ( ) **2 ( ) **2 40 36 40 48 12 − + − = Euclidean Distance between Point A and training field C ( ) **2 ( ) **2 40 44 40 41 5 − + − = Round the Block Distance Distance to Point A = |(40-36)+(40-48)| |4 + -8 | = 12 Distance to Point C = |(40-44)+(40-41)| | -4 + -1 | =...
View Full Document

This note was uploaded on 12/21/2010 for the course GIS 4037c taught by Professor Roberts during the Spring '10 term at FAU.

Page1 / 8

9MinDistPowerpnts - – Many clusters, small cluster...

This preview shows document pages 1 - 8. Sign up to view the full document.

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