Nature of an Image Flashcards
digital RS imagery
- array of scene elements “pixels”
- based on raster format
- not a photograph
- a pixel records EMR from Earth as a digital number
Image
a rendition of target features described through the use of EMR
sensors recording EMR determine ____
resolution
Tpes of resolution
- spatial
- spectral
- temporal
- radiometric
Pixel location and size:
ground footprint
pixel brightness is controlled by:
nature of the target
allows creation of spectral signature
levels of pixel brightness determined by:
radiometric resolution (bit depth)
types of imaging systems
- frame
- scanner
- whiskbroom
- pushbroom
visible light wavelength range
400-700 nm
Spectral resolution
number of spectral bands
spatial resolution
pixel size or scale
temporal resolution
timescale
Multispectral Image
image composed of ‘n’ rows and columns of pixels for multiple spectral bands
more than one raster data set or spectral band
what is each data set in a multispectral image
a spectral band
multispectral images detect ____ in multiple bands of EMR
energy
how to display a colour composite with a multispectral image
three bands combine into red, green, and blue
Binary Numbers
Bit depth (2^n), represent decimal numbers
.# of bits = # of decimal levels
how are digital numbers typically stored
binary form
Digital numbers are a type of what kind of resolution and why
radiometric, because the binary number (BIT) is also the amount of grey levels)
additive colour
RGB on top of each other create white light, and selectively adding RGB creates YMandC
Subtractive colour
YMC superimposed create black, with RGB in the overlap margins
true colour composite
RGB guns match RGB bands
why digitally process RS data
- capable of more detail and precision than humans
- humaninterpretation is subjective and not always repeatable
- computer can better handle large amounts of data
- can manipulate images with low contrast
digital number
numerical digit assigned to a pixel
compute image histogram
.# of pixels vs DN selected from a region of an actual image
DN in a histogram
brightness value represented as anumber (lowest is darkest)
why enhance
Increase contrast
histogram analysis
examining distribution of pixel values in an image to understand characteristics
linear contrast enhancement
min-max stretch
stretch entire histogram linearly based on min and max values
percentage contrast enhancement
saturation stretch
zero in on portion of histogram we are interested in and give whole dynamic range to that part
pros and cons of percentage stretch
lose variability above and below threshold but gain max contrast in our section
outside of our threshold the image is black or white, fully saturated
image composite interpretation
composite images by combining multiple bands or channels to enhance specific features
