EX1-1 Flashcards
Active remote sensing
An active sensor system provides its own energy source. As an example, a radar sensor sends out radio waves and records the reflection waves coming back from the surface.
Radar; Lidar
Passive remote sensing
Passive systems are much more common than active systems
• A passive sensor system needs an external
energy source. In most cases this source is
the sun.
• These sensors generally detect reflected
and emitted energy wave lengths from a
phenomenon
Optical; air photos; multispectral; hyperspectral
Spatial Resolution
It is the size of the smallest area that can be separately recorded as an entity on the image.
Spatial resolution depends on:
Property of sensor (e.g. Focal length of camera)
Altitude of sensor (e.g. platform height and stability)
Temporal Resolution
minimum time period between two images exact same area same viewing angle same platforms using the same instruments comparable conditions
weather satellites have high temporal but low spatial resolution.
Spectral Resolution
observed spectral differences in the energy (wavelength) reflected or emitted from features of interest.
the ability of a sensor to detect small differences in wavelength. The narrower the wavelength range, the finer the spectral resolution.
Radiometric Resolution
sensitivity of the sensor to incoming radiance or the differences in the brightness of objects.
(How much change in radiance is required to result in a change in recorded brightness value?)
Electromagnetic Radiation (principal spectrums and those that are useful for remote sensing)
Panchromatic – 1 Band (B&W)
Color – 3 Bands (RGB)
Multispectral – 4+ Bands (RGBNIR)
Hyperspectral – 100s of Bands
Limits of the visible spectrum are defined by the sensitivity of the human visual system. It can be divided into 3 segments (“additive primaries”) – B, G and R
Very large spectrum; Categorized into NIR, MIR and Far IR (also known as thermal IR). Along with visible spectrum, NIR and MIR form the “reflective spectrum” (i.e., they are essentially solar radiation reflected from the Earth’s surface). Far IR / thermal IR is radiation emitted by the Earth.
Relationship between frequency, wavelength, energy and speed of light
speed of light - c
Frequency - ν
Wavelength – λ
c = λ ν
Frequency and wavelength are inversely proportional
Emissivity
ratio between the emittance of a given object and that of a blackbody at the same temperature
•Emissivity ranges from 0-1 and describes an object’s
ability to behave like a black body
Blackbody
A blackbody radiator is a hypothetical body
that absorbs and re-radiates all energy
incident upon it with no change in energy
A blackbody emits energy with perfect
efficiency
Whitebody
A whitebody is a hypothetical body that
reflects all energy incident upon it with no
absorption or re-radiation of energy.
Graybody
Many bodies have a constant emissivity regarding the wave length, but do emit far less radiation than black bodies. They are called grey bodies.
Selective radiator
Bodies whos emissivity depends on the temperature and the wave length, such as metals, are called selective radiator.
Kirchoff’s Law (understanding of law and its implications for remote sensing)
DEFINES BLACKBODIES:
Ratio of emitted radiation/absorbed radiation
is the same for all blackbodies at the
same temperature.
This law forms the basis for the definition of emissivity.
The emissivity of an object varies at different wavelengths. efficiency of absorption and emission of radiation are material properties.
Stefan – Boltzmann Law (understanding of law, the formula and its implications for remote sensing)
This law defines the relationship
between the total emitted radiation (W)
(often expressed in watts per unit area)
and temperature
Stefan - Boltzmann Law: Total emission radiated from a blackbody is proportional to the fourth power of its absolute temperature (T): W = σT4,
where: W is the total emission radiatied
T is the absolute temperature in (K)
σ is the Stefan – Boltzmann constant
(5.6697 X 10-8)
Hot blackbodies emit more
energy per unit area than do cool
blackbodies
Wien’s Displacement Law (understanding of law, the formula and its implications for remote sensing)
Wien’s Law: The wavelength of max radiation is a function of an object’s temperature (T): λ max = 2898 mm / T, where: λ max is the wavelength at which radiance is at a maximum T is the absolute temperature in (K) This law specifies the relationship between the wavelength of radiation emitted and the temperature of a blackbody As blackbodies become hotter, the wavelength of maximum emittance shifts to shorter wavelengths
Reflectance and emittance of electromagnetic radiation
-
Atmospheric interactions of EMR – transmission, scattering, absorption
Scattered (i.e. reflected or refracted by atmospheric particles) before it reaches the
Earth’s surface
• Absorbed by atmospheric particles before it reaches Earth’s surface
• Transmitted to the earth’s surface, where it is either reflected, transmitted or
absorbed
• Absorbed by an object on the Earth’s surface, and then emitted by that object
transmission (Atmospheric)
• Transmission is what we want the atmosphere to do when we are studying
the earth itself.
• Desirable, but even under the best of conditions, 100% transmission does
not occur
• This is because some radiation is scattered, absorbed or reflected by
atmospheric gases
scattering (Atmospheric)
- Scattering is the redirection of radiation by particles suspendedin the atmosphere or by large molecules of atmospheric gases.
The mount of scattering depends on:
Size of atmospheric particles (from molecules of gases to dust particles to larger ice and water droplets)
The wavelength of radiation
absorption (Atmospheric)
Absorption occurs when the atmosphere prevents, or strongly attenuates, transmission of radiation through the atmosphere. Characteristics of absorption: Energy absorbed by the atmosphere is subsequently reradiated / emitted at longer wavelengths Different gases absorb well at different λ O3 absorbs …… CO2 absorbs……. WV absorbs …….(role of WV varies greatly with time and location)
3 kinds of scattering
Rayleigh, Mie, and Non-selective
Rayleigh scattering
Occurs when λ (shorter) > particle size;
Occurs primarily due to molecules of atmospheric gases;
Scattering inversely proportional to fourth power of λ
A perfectly clean atmosphere, consisting only of atmospheric gases, causes
scattering such that the amount of scattering increases greatly as λ becomes
shorter
Violet light scattered 16X Red light; Blue light scattered 4X Red light
Mie scattering
Occurs when λ (longer) = particle size;
Occurs primarily due to dust, pollen, smoke, water droplets and
cloud ice particles (all of which are particles larger than the
atmospheric gas molecules responsible for Rayleigh scattering)
Tends to affect longer λ in and near the visible spectrum
Non-selective scattering
Occurs when atmosphere is heavily dust laden
Scattering particles include larger water droplets or large particles of airborne dust
“Non-Selective” scattering not λ dependent all visible λ scattered equally appears as whitish or
grayish haze
Absorption and atmospheric windows
Those wavelengths that are not absorbed and are relatively easily transmitted through the atmosphere are called “atmospheric windows”
Major windows (table 2.4 textbk):
UV and visible
NIR
Thermal
Microwave
We can see in precisely the portion of the spectrum that is not absorbed by the atmosphere (evolution and adaptation of our eyes!!)
Atmospheric windows define those wavelengths that can be used to form images and are therefore effective for RS
Surface interactions of EMR
reflection, absorption and transmission
reflection (Surface)
Surface roughness < incoming λ = specular reflectance
Surface roughness > incoming λ = diffuse reflectance
Most surfaces are somewhere between specular and diffuse
Important Laws related to Lambertian surface reflectance
Lambert’s Cosine Law: The perceived brightness (radiance) of a
perfectly diffuse surface does not change with the angle of view
Inverse Square Law: The observed brightness decreases according
to the square of the distance from the observer to the source
absorption (Surface)
• The proportion of energy reflected, absorbed or
transmitted varies for different earth features.
• Within a given feature, the proportion reflected, absorbed
or transmitted will vary with different λ
transmission (Surface)
• The proportion of energy reflected, absorbed or
transmitted varies for different earth features.
• Within a given feature, the proportion reflected, absorbed
or transmitted will vary with different λ
2 kinds of reflection
Surface roughness < incoming λ = specular reflectance
Surface roughness > incoming λ = diffuse reflectance
specular reflection
Angle of incidence = angle of reflection;
reflectors are flat, mirror like surfaces (mirror, smooth metal, calm water body) that redirect radiation in a single direction
The surface is smooth relative to the λ
diffuse reflection
reflectors are rough surfaces that reflect uniformly in all directions
The surface is rough relative to the λ
A perfectly diffuse reflector (Lambertian surface) would have equal brightness when observed from any angle.
Atmospheric correction
•Reduces the effects of haze and absorption
• Improves the sharpness of images
• Allows normalization of images
– Provides for better comparison between different images
Spectral properties of objects and how remote sensing uses these properties to distinguish between objects
• The spectral signature of an object is its own unique
spectral response pattern (much like a fingerprint) and
helps distinguish the object from other objects
• It is a record of the unique reflectance properties of an object at different wavelengths
Spectral signatures (while you don’t need to memorize spectral signatures of all objects, you should have a general idea of how certain common objects such as water and vegetation behave). You should be able to interpret spectral signature curves and understand the information that they convey
• The proportion of energy reflected, absorbed or
transmitted varies for different earth features.
• Within a given feature, the proportion reflected, absorbed
or transmitted will vary at different λs
• Thus, 2 features may be indistinguishable in one
spectral range, but they may be very different in another.
• Spectral reflectance (the proportion of incident energy
that is reflected) of an object can be plotted for different
wavelengths to produce a spectral reflectance curve
• Comparison of spectral reference curves of different
objects can help us figure out the most useful bands
for distinguishing between these objects
BW Panchromatic imagery
- Panchromatic means “across the colors”
- The entire visible spectrum is represented as a single channel / band without distinguishing between R, G and B
- This provides a B&W image that records brightness using radiation from the visible spectrum without separating the different colors (a regular B&W photo)
- Idea is to use data capacity to capture spatial detail rather than a colored representation
- Because Blue light scatters easily and degrades image quality, some instruments capture radiation across G, R and IR regions of the spectrum, thereby providing a sharper, clearer image.
- Traditionally, panchromatic referred only to the visible spectrum, but now includes a broader region extending into the NIR
Natural Color Imagery (or True Color Composite)
- Natural color imagery is how our eyes / visual system applies band combinations
- Totally obvious – we see Blue as Blue, Green as Green and Red as Red (a normal color photo)
- It is a familiar representation of scene, but a disadvantage in remote sensing is that ……..
False Color Imagery (or False Color Composite)
Our eyes can perceive only the visible spectrum.
How can we record / represent radiation from outside the visible spectrum?
When recording radiation from outside the visible spectrum, we must make assignments that depart from the natural color model
This creates false color images (things don’t appear the color that we naturally perceive them to be)
Seems nonsensical to represent objects using color other than their natural colors
But an essential task in RS
False color assignment is arbitrary – but certain assignments are commonly accepted
Color Infrared Imagery
NIR very useful for RS
Use of NIR adds a fourth spectrum channel to the
natural color model
Because we recognize only three primaries, adding
an NIR channel requires omission of one of the
visible bands
Which band is easiest to exclude??? Why?
CIR model creates a 3 band color image by excluding … and including NIR
Initially developed in WW II for camouflage detection (to differentiate between actual vegetation and surfaces painted green to resemble vegetation)
Pixel values / Brightness values / Digital numbers
Ls = (Gain x DN) + offset (or “bias”) DN = (Ls – offset)/Gain
Digital numbers and binary numbers (you need to understand the concept but you won’t get any quantitative problems requiring conversion from one to the other)
-
Bits, Bytes, Kilobytes, Megabytes, Gigabytes
1 byte = 8 bits
1 kilobyte = 1024 bytes (= 2^10)
1 megabyte = 1,048,576 bytes (= 2^20)
1 gigabyte = 1,024 megabytes (= 2^30
How to calculate radiometric resolution of a raster image if the range of DN values are known
know your base 2 calculations
How to calculate file size of a raster image if number of rows, number of columns, number of bands and bit depth are given
L X W X BITDEPTH/8
Image enhancement
• Image enhancement is the process of improving the
visual appearance of digital images
• An arbitrary exercise
• Integrity of original data altered
• original BVs are altered in order to improve visual qualities
• BVs lose their relationship to original brightness on the ground
• Therefore, for further analysis, original and not enhanced
image values should be used as input
Common Image enhancement techniques
(Linear stretch, Histogram Equalization, Density Slicing, Edge Enhancement)
Linear stretch
Linear stretch converts the original digital values into a new distribution, using new min and max values.
This stretch simply takes a range of the data and rescales it to a preset range. For example, 50 to 190 becomes 0 to 255
This technique is useful for datasets that are concentrated in the middle of the grey scale
Since it simply rescales data, it is useless if there is even 1 pixel at either end of the greyscale. In this case, the rescaling has no effect on the data
Histogram Equalization,
Histogram Equalization is a stretch that favorably expands some parts of the DN range at the expense of others by dividing the histogram into classes containing
equal numbers of pixels.
For instance, if most of the radiance variation occurs in the lower range of brightness, those DN values may be selectively extended in greater proportion to
higher (brighter) values
**Pixels at peak are spread apart -> Contrast gained
Pixels at tail are grouped -> Contrast lost
Density Slicing
Density slicing is accomplished by arbitrarily dividing the range of brightness in a single band into intervals, then assigning each interval a color.
Density slicing emphasizes certain features that may be represented in vivid colors, but it does
not convey any more information than does the single image used as the source.
Edge Enhancement
• Techniques discussed so far involve moving pixels around on the grey
scale.
• Each operation affects all pixels with the same grey scale in exactly the
same way, without regard for context
• Edge enhancement is different.
• Edge enhancement makes use of spatial filters to enhance boundaries /
transitions between features
• What are spatial filters?
The concept of spatial filters, how they work and how they change images
(low pass filters, high pass filters, edge detection / enhancement filters)
low pass filters
Low pass filters are called smoothers because they
suppress detail, and emphasize continuous tones
They create regions of homogeneous BVs
A “mean” filter is a low pass filter.
It multiplies the data in the window by 1, sums the result, and divides the
summation by 9 (the total # of pixels in the filter)
The result (which is the mean) is placed in the central pixel of the output
Notice the output the magnitude of change is reduced!
high pass filters
High pass filters enhance detail by emphasizing frequent
changes in BVs as the filter moves across the image
Useful for urban images
edge detection / enhancement filters
Edge detectors enhance tone changes between two pixels
Used to sharpen roads, field boundaries, other linear
features
Image scale and distance measurements (map scale, distance on map and distance on ground – how to calculate one unknown when 2 knowns are given)
DISTANCE BETWEEN A & B is 4.6 inches:
MAP SCALE = 1:25,000
1/25,000 = MD/GD
MD = 4.6 GD = 4.6* 25,000 = 115,000 inches
NOW: We divide 115,000 by #of inches in a mile.
115,000/63,360 = 1.82 Miles
_____________________________________
PD/GD = 2.2 inches/115,000 inches = 1/52,273
Concept of remote sensing
• Remote sensing is the science and art of obtaining
information about an object, area, or phenomenon through the analysis of data acquired by a device (sensor) that is not in contact with the object, area or phenomenon under investigation
• Because there is no direct contact between the sensor and object of study, some remote interaction must be introduced to carry information from the object to the sensor.
• The interaction between electromagnetic radiation and the object is the most common interaction used in modern remote sensing
definition of remote sensing
Remote sensing is the practice of deriving information
about the Earth’s land and water surfaces using images
acquired from an overhead perspective, using
electromagnetic radiation in one or more regions of the
electromagnetic spectrum, reflected or emitted from the Earth’s surface.
overview of remote sensing
• Physical Object:1) Objects that need to be examined
• Sensor Data: 2) The instrument (camera, radar) that
views the physical objects by recording electromagnetic radiation emitted or reflected from the object
• Extracted Information: 3) Transformation / analysis /
interpretation of sensor data (which is often abstract due to perspective, resolution and use of spectral signatures outside the visible spectrum)
• Applications: 4) The analyzed remote sensing data is
combined with other data to address a specific practical problem
