Note:- Please prepare all these 4 Question as it will be asked in the class test on 16th August i.e Thursday, these questions + material given below is IMP for the MSE-I Examination
1) What is spectral signature?
Ans:-According to spectral reflectance curve different objects have different reflectivity at different wavelengths. It causes any object to form a pattern on spectral reflectance curve which is unique. This pattern can be referred to as the spectral signature of the object. In other words it can be said that the measurements taken at different wavelength of different objects shows their spectral signature.
At certain point on the spectral signature the object will appear the most distinct therefore it helps to recognise particular object and also allow the observer to select the most preferable wavelength for the object. Take an example of water and vegetation; both will appear same in visible spectrum but there is a considerable difference between them in near infrared range. Therefore if the observer is interested in observing vegetation, infrared wavelength will be preferred as it will allow the most distinct appearance of vegetation which will appear bright due to higher reflectivity in infrared region.
2) Explain image enhancement ?
Ans:- Image enhancement
•
Image reduction
The images that cannot be viewed on the screen due to their size are reduced in size by image reduction technique. In this the number of pixels are reduced in such a way that the image is reduced to 50% or 25%. However while doing so important data may be lost as rows and columns of alternating pixels are to be removed.
•
Image magnification
It is opposite to the image reduction as here the image is zoomed by increasing the number of pixels in the image. To make the image twice bigger additional three pixels of same digital values are added to the original pixel.
•
Colour compositing
There are three basic colours blue, green and red. Different colours can be obtained through addition of proportionate mixture of these three colours. This method is called additive colour composite. If three pigments of primary colours i.e. cyan, magenta and yellow are used than different colours are obtained through subtraction and therefore this method is known as subtraction colour composite.
If the obtained image has waves from Infrared region than they are converted into visible colours which means the object appears in completely different colour than the real world situation. This is known as false colour composite. If the image is in visible colour region then it is known as true colour composite.
•
Transect Extraction
In some images there is a clear distinction between two regions. It suggests that some images do not require analysis with as many bands as they did. Therefore transect extraction is carried out in order to determine which band is giving the highest differences in reflection profile of a particular image.
•
Contrast enhancement
There are two different types through which contrast enhancement can be carried out...
Linear stretch
The shades are increased in order to increase the number of distinctive objects in the image. As the name suggests the image is stretched.
Non linear stretch
It is good for low contrast image and there are two main procedures through which non linear stretch can be carried out.
Histogram equalisation
With the help of histogram all objects are specified in different shades according to their brightness value. This method greatly helps in differentiating different shaded regions on the basis of their histogram.
Gaussian Stretch
Linear starch method and Histogram provides different shades for a given image but they greatly reduce the distinctive appearance of highest and lowest bright contrasts in the image. Gaussian stretch greatly improves the distinctive quality of highest and lowest bright contrasts. On the other hand it reduces the differential contrast in the middle region.
•
Density slicing
It converts number of shades (between black and white) into different regions of colours so that human eye can easily extract information from the image which instead would have appeared as non recognisable shades. Sometimes symbols are also given instead of colours.
•
Spatial filtering
It is known that the image has lower and higher frequencies depending upon the reflectance profile of different objects. If image has objects having almost same spectral signature than it is said to have low frequency while on the other hand if the image contains objects having distinct spectral signature than the frequency is considered as higher.
Depending upon the requirement the frequencies are required to be higher or lower in terms of frequency. In order have such images filtering is carried out which may be low pass or high pass filtering.
Low pass filters remove high frequency brightness values and are known as spatial smoothing filters. High pass filters remove low frequency brightness values and are known as spatial sharpening filters.
•
Edge detection
Some natural features such as geographical faults or the river passing by are clearly distinguishable due to abrupt colour difference on the image. Same way there are some manmade that features that can be separated from the background easily such as roads, canals, railway lines etc.
If in some case this type of contrast doesn’t appear than it has to be enhanced by edge enhancement techniques. Two edge enhancement techniques are discussed below,
Directional filters: These filters enhance the linear edges in horizontal, vertical and diagonal direction.
Non directional filters: These filters can enhance linear filters as well as non linear filters by Laplacian filters.
Ques3) What is image classification? Explain supervised and unsupervised classification.
Ans:- Image classification
•
Supervised classification
As the name suggests supervised classification consist of data entry from the analyst. It provides information about several features such as forest areas, dry sand, and water to the computer by allocating sample regions. These sample regions help the computer in determining DNs for each region. It then autonomously classifies all regions based on their brightness values.
There are different algorithms to perform the autonomous classification such that minimum distance to means and maximum likely hood.
•
Unsupervised classification
Unsupervised classification has computer deciding everything i.e. from the initial information to the final classification. The user remains unaware of the method based on which the classification is carried out. The computer uses its own database to allocate DNs to the features in the image.
Unsupervised classification is very useful in knowing the spectral classes into which the pixel fall while it has to be defined by user in case of supervised classification. However the data provided by the unsupervised classification has to be analysed further for precise outcome by the supervised classification.
Ques4) Explain RADAR.?
Ans:- Radar uses microwave as a medium to sense objects that cannot be or hard to be analysed with visible spectrum. They provide distinct information about the objects and many applications exist due to their visibility and penetration ability.
Radar sends and receives continuous pulses in order to collect information in terms of the objects position relative to the radar’s position. It also provides details such as size and shape of the objects.
Radar mainly consists of the following components,
•
Pulse generator discharges times pulses
•
Transmitter generates successive short burst at regular interval
•
Duplexer co ordinates active transmitted & received microwave energy
•
Directional antenna shapes and focuses pulses into stream of pulses
•
Receiver antenna receives returned pulses and sends it to receiver
•
Receiver converts pulses and sends it to receiver
•
Recording device stores the information in digital mode for later processing in digital tape recorder or hard disk
•
Cathode ray tube monitor produces real time display
Applications of radar
•
Static radars are used on airport for the air traffic control
•
Navy ships have radars to sense disturbances and intrusion
•
Air borne radars are used for taking microwave images of ground bases areas
•
Possible infiltration in air space is effectively encountered by surveillance through radars in air crafts
•
Space borne radars are used to gather information about planet’s atmosphere
•
Information about metal’s ores beneath the earth soil can be easily obtained using radar
Types of radar
•
PPI (plan position indicator radar)
o
This type of radar has circular display screen on which continuous radar pulses creates echoes of the objects
o
Its spatial resolution is lower therefore it is used in weather broadcasting, air traffic controls and navigation purposes
•
SLR (side looking radar)
o
Side looking radar has a side pointed antenna that directs the transmitting pulses; the antenna is generally placed below the aircraft or spacecraft
o
The radar produces continuous side strips that cover huge ground areas parallel to the air borne or space borne vehicle.
o
If SLR is on air craft than it is called SLAR (side looking airborne radar)
********************************************************************************************
For Reading Purpose........
Image statistics
and sub setting
A digital image
contains pixels having different digital numbers. Each number represents
specific smallest portion of the image. Therefore these digital numbers in
combination provides a complete view of the original image. Allocation of
digital numbers to the image is carried out in such a way that specific
information can be extracted easily.
It is quite
obvious that more number of smaller pixels creates better image as smallest
features can be included with decreasing pixel size. However this type of image
occupies more space. On the other hand if bigger pixels are used than they
occupy least amount of space but the fineness of the image has to be
compromised.
Number of bands
used to capture the image also matters while discussing about image statistics
as they play a major role in deciding the quality and information content
within the image. If a single band is selected for detection by sensor than it
will show all the rest of colours in changing shade It is to be noted that in
this case the detailing depends upon the number of pixels allocated to the
image. Such image which is known as panchromatic image however is not
considered as a detailed and easily understandable image due to its single
colour display.
Multispectral
image on the other hand has multiple bands of different colour regions such
that combination of all bands give real image in which all the colours can be
recognised in easy and similar manner to the actual human eye view.
Sub setting
Sub setting
means breaking of the bigger image into smaller parts. When the analyst is
interested in only few objects or features from the image than the images is
broken into several parts using the function known as sub setting. There are
two types of sub setting i.e. spatial sub setting and spectral sub setting.
Spatial sub
setting is breaking of the image into several parts by selecting specific
region of pixels from the image.
Spectral sub
setting is selecting particular wavelength from the multispectral image to
acquire information about specific band. Thus this image converts multispectral
image into lesser amount of bands or into panchromatic image.
***********************
For Reading Purpose........
Digital
interpretation of true colour and false colour composite
The number of
colours that can be represented by the monitor is known as colour resolution.
As in case of radiometric resolution the numbers that can be obtained depend
upon the bits of sensor. 4 bit sensor will give 24=16 different
shades in black and white monitor.
Colour monitor
has these bits shared between three colours equally. Therefore if 24 bit sensor
is used than 8 bits are allotted for each colour and therefore 28 shade
for each colour can be obtained. Suppose 6 different bands are used than
obtained colours will be 6! / (6-3)! = (1.2.3.4.5.6)/ (1.2.3) = 120.
Colourful image
consists of basically three main colours, i.e. blue, green and red. It’s the
vastly different proportion of these colours that provides many different
colours. Its major advantage is the obtained similarity of the captured image
in terms of colours with the one that humans are used to see with their eye.
There are two
different colour composite types that fairly explain all colourful images that
are captured through remote sensing technique,
1)
True colour composite
2)
False Colour composite
True colour
composite
As the name
suggests this type of image is made up of three main different colours
therefore the image is almost similar in colour to the actual vision of human
eye. Such resemblance removes many complexities while understanding the
features of the image.
False colour
composite
For the image
having false colour composites the actual colour of the image doesn’t appear in
the image instead there is a considerable difference compared to the actual one
that is noticed through human eye on earth.
If one of the
band in multi spectral image is in region apart from visible one than it
becomes necessary to allot a colour to that invisible band so that difference
between various objects in the image can be identified by the analyst.
One of the most
common colour composite schemes is shown below,
R=XS3 (NIR band)
G=XS2 (Red band)
B=XS1 (Blue
band)
where red colour
is allotted to the NIR band while green colour is for red band and consequently
blue colour for green band.
In similar way
R= SWIR band
G=NIR band
B=Red band.
*********************
For Reading Purpose........
Visual Image Interpretation
Image interpretation process includes image reading, image measurement and image analysis.
• Image Reading: - It is the most basic form of image interpretation. Image reading is carried out by properties of the objects in the image, i.e. shape, size, pattern, tone, texture, colour, shadow etc.
• Image measurement: - As the name suggests it provides information about dimensional features such as length, location, height, density, temperature.
• Image analysis: - It is execution of information from the image through formerly obtained data through image reading and image measurement.
The information obtained through such way is then used to create more understandable map which is known as interpretation map or thematic map.
Elements of visual image interpretation
• X,Y location
• Size
• Shape
• Shadow
• Tone/colour
• Texture
• Pattern
• Height/ depth
• Site/situation/association
X, Y location: - There are two basic methods through which important information can be obtained; I) Surveying in the field and II) Collecting remote sensing data
Size: - It provides information about length, width, area, perimeter etc about the object.
Shape: - It provides information about shape of the object such as an airport can be visualized as a straight strip from the space while the football stadium will appear as the large oval shaped object.
Shadow: - shadow of the object sometimes gives more information compare to the object itself. Especially height can be obtained from the map. If the time is known than the angle made by the incident sunrays to the surface is used to find out the height. On the other hand if the height is known than the time of the image taken can be found out.
Tone: - In temporal resolution the shade differs from black to white. This depends upon the inherent characteristics of the object. The object at higher temperature will emit high amount of radiation compared to the object with the lower temperature.
Colour: - The vegetation will appear as green if EM waves in visible range are sent towards the object because the object will reflect green colour and rest is absorbed by the leaves of vegetation.
Texture: - Texture can be defined as placement and arrangement of repetitions of colour or tone in the image. Smaller individual objects can not be identified easily. In this case texture turned out to be useful feature to find out presence of smaller objects as a group.
Pattern: - As the name suggests the patterns are useful in determining specific areas which differs from others. Take an example of comparison of urban areas and areas of outskirts. Urban areas have pattern of house rows and roads in denser manner compared to the outskirts.
Height and depth: - Height and depth of objects can be found out using parallax which depends upon the displacement of the object due to change in observation point.
Site situation and association: - It depends upon the understanding and intuitive nature of the humans. Sometimes there are obvious places on the map which suggests particular site to be existed. Such as coal storage and cooling pond nearer to the thermal power station.