final

Question 1

What is an arbitrary or spatial profile?

Answer 1

graph across space of spectral profile (e.g. transect from inland marsh to middle of river)

Question 2

Spectral profile

Answer 2

looks at individual point, or spectral signature for that pixel

Question 3

Convolution filtering

Answer 3

window that moves over an image sequence and does a function on the pixels it encounters

Question 4

NDVI

Answer 4

NIR-red / NIR + red

Question 5

histogram matching

Answer 5

histogram matching tries match responses of classes

Question 6

linear contrast enhancement or contrast stretching

Answer 6

expands the original digital values of the remotely sensed data into a new distribution

Question 7

what would a filter look like that accentuates spacial variability?

Answer 7

a linear edge detector would have some negative and some positive values

Question 8

principal component analysis

Answer 8

allows you take multiple bands, take the variability and store it in just a few bands. Output number of bands is the same but the crucial information is contained in the first few bands. Thus the benefit is that you can decrease the data dimensionality (number of bands)

Question 9

data dimensionality

Answer 9

number of bands, size of your data defined by rows and columns.

Question 10

“features” in image processing (different in GIS)

Answer 10

bands

Question 11

What is the point of features selection

Answer 11

another way to get rid of data you don’t need. It’s determining which bands are the most important

Question 12

NDVI stands for

Answer 12

Normalized Difference Vegetation Index

Question 13

Tassled cap transformation

Answer 13

the first of the new axis becomes brightness (formed from soil line). the axis orthogonal to it is called greenness. Then wetness is sometimes the third axis

Question 14

supervised vs. unsupervised

Answer 14

unsup clusters elements in the image that look similar. supervised always involves a training data set that you give the algorithm

Question 15

cluster busing

Answer 15

something that is done in unsupervised classification. we take the classes that didn’t get classified and reclassify them

Question 16

Why is it a good idea to used predefined classification schemes

Answer 16

a class scheme like USGA or Anderson is useful because it allows us to monitor the change over time if we have a classification scheme that stays consistent. also involves less work.

Question 17

problems with extending spectral signatures through space and time

Answer 17

you’re telling the algorithm that forests look like this..that may work fine in one single image but the atmosphere information in different. sun was different.

Question 18

feature space

Answer 18

a graphical representation of the pixels by plotting 2 bands vs. each other (often looks like a tasseled cap)

Question 19

pros of parallelepiped

Answer 19

simple, makes a few assumptions about the character of the classes

Question 20

cons of parallelpiped

Answer 20

parallelelpipeds are rectangular, but spectral space is “diagonal” so classes may overlap

Question 21

minimum distance to mean

Answer 21

finds mean value of pixels of training sets in n-dimensional sapce. All pixels in image classified according to the class mean to which they are closest.

Question 22

pros of minimum distance

Answer 22

all regions of n-dimensional space are classified. Allows for diagonal boundaries (and hence no overlap of classes)

Question 23

cons of minimum distance

Answer 23

assumes that spectral variability is same in all directions, which is not the case

Question 24

maximum likelihood

Answer 24

assume multivariate normal distributions of pixels within classes. calculates the probability that the pixel is a member of that class. takes into account mean and covariance of training set. each pixel is assigned to the class for which it has the highest probability of membership

Question 25

pros of max like

Answer 25

most sophisticated; achieves good separation of classes

Question 26

cons of max like

Answer 26

requires strong training set to accurately describe mean and covariance structure of classes

Question 27

how do you choose locations fro training data?

Answer 27

make sure it’s representative spectrally of the classes you have. it may be that you need to come up with multiple types of forest if they are significantly different in their response.

Question 28

soft classification logic

Answer 28

outputs a probability that something is in a class

Question 29

ISODATA pros

Answer 29

very efficient at identifying spectral clusters within data

Question 30

how ISODATA works

Answer 30

algorithm that splits and merges clusters; user dines threshold values for parameters; computer runs algorithm through many iterations until threshold is reached.

Question 31

ISODATA uses _____ method

Answer 31

shortest distance to center

Question 32

(ISODATA) the _______ within each cluster and the distance between cluster centers is caluculated.

Answer 32

SD

Question 33

clusters are _____ if one or more SD is greater than the user defined threshold. clusters are ____ if the distance between them is less that the user-defined threshold.

Answer 33

split; merged

Question 34

expert system

Answer 34

tries to emulate the rules that a human expert would use

Question 35

knowledge base

Answer 35

the extraction of expert’s rules

Question 36

inference engine

Answer 36

uses rules of expert system

Question 37

5 change detection algorithms

Answer 37

Write Function Memory Insertion
image algebra (differencing, ratioing)
post classification comparison

Question 38

spectral angle difference

Answer 38

uses n-D angle to match pixels to reference spectra. the algorithm determines the spectral similarity between two spectra by calculating the angle between the spectra and treating them as vectors in a space with dimensionality equal to the number of bands.

Question 39

the only change detection algorithm that results in a to-from contingency table?

Answer 39

post classification change detection

Question 40

user’s accuracy is based on ____

Answer 40

the classification data (divide the number in the diagonal by the total classified in that catagory)