Most of the common image processing functions available in image analysis systems can be categorized into the following four categories:
1. Preprocessing (Image rectification and restoration)
2. Image Enhancement
3. Image Classification and Analysis
4. Data Merging and GIS Interpretation
It is the task of processing and analyzing the digital data using some image processing algorithm. The analysis of relies only upon multispectral characteristic of the feature represented in the form of tone and color.
Digital image processing (DIP)
UNIT III
Techniques of Digital Image ProcessingInitial Data Statistics
Statistical information such as minimum and maximum values of the data set, mean, standard deviation, and variance for each band are calculated. Histograms and scatter-grams provide a graphical view of the nature of different bands.
Image Rectification and Restoration ( or Preprocessing)
These are correction needed for the distortion or degradations of raw data. Radiometric and geometric correction are applicable to this.
Image Enhancement
Purpose of this is to improve the appearance of the imaginary and to assist in subsequent visual interpretation and analysis. Normally, image enhancement involves techniques for increasing the visual distinction between features by improving tonal distinction between various features in a sene using technique of contrast stretching.
Image Transformation
These are operations similar in concept to image enhancement. Generally, image enhancement operation is carried out on a single band of data, while image transformations are usually on multiple bands.
Image Classification
The objective of the classification is to replace visual analysis of the image data with quantitative techniques for automating the identification of features in a scene.
Digital Data
Initial Statistic ExtractionInitial Display of Image
Ancillary Data
Image Enhancement Visual Analysis
Image Classification
Unsupervised Supervised
Classified Output
Post processing operation
Data Merging
Assessment of Accuracy
Report DataMaps and Images
Image Rectification and Restoration
Image Enhancement (Histogram Example)
Original
Histogram Example (cont. )
Poor contrast
Histogram Example (cont. )
Poor contrast
Histogram Example (cont. )
Enhanced contrast
Image HistogramAn image histogram is a
graphical representation of the
brightness values that
comprise an image. The
brightness values (i.e. 0-255)
are displayed along the x-axis
of the graph and the frequency
of occurrence of each of these
values in the image on the Y-
axis. By manipulating the
range of digital values in an
image, i.e. graphically
represented by its histogram,
various enhancement can be
applied to the data. However,
these can be grouped under
two categories:
1. Linear contrast Enhancement
2. Non linear contrast Enhancement
Spatial FilteringSpatial filtering is the digital processing function that are used to enhance the appearance of an image. Spatial features are designed to highlight or suppress specific features in an image based on their spatial frequency.
Spatial frequency is related to the concept of image texture. It refers to the frequency of the variations in tone that appear in an image. Rough texture areas of an image, where the changes in tone are abrupt over a small area, have high spatial frequencies, while smooth areas with little variation in tone over several pixels, have low spatial frequencies.
Types of Filters
1. Low-pass Filter: is designed to emphasize large homogenous areas of similar tone and reduce the smaller detail in an image. Thus, these filters generally serve to smooth the appearance of an image.
2. High-pass Filter: such filters do the opposite job as low-pass filter. They are served to sharpen the appearance of fine details in an image.
Other Filters
1. High boost filters
2. Directional or edge detection filters
Smoothing and Sharpening Examples
Smoothing(Low-pass Filter)
Sharpening(High-pass Filter)
Image Classification
• To identify and map areas with similar characteristics
• To assign meaningful categories such as land-use or land-cover classes to pixel values
Purpose
1. Supervised classification
2. Unsupervised classification
Classification Methods
Supervised Classification
In this classification method, an analyst identifies the imaginary in terms of homogenous representative samples of different surface cover type of interest. These samples are called as “Training Areas”.
The selection of appropriate training area is based on the analyst’s familiarity with geographical area and knowledge of the actual surface cover types present in the image.
The numerical information in all spectral bands for the pixels comprising these areas are used to train the computer to recognize specially similar areas for each class.
Therefore, in supervised classification, the analyst is first identifies the information classes based on which it determines the spectral classes which represent them.
Common Classifiers: 1. Parallel-piped classifiers
2. Minimum distance to mean classifiers
3. Maximum likelihood classifiers (MLC)
Digital Image
Supervised Classification
The computer then creates...
Supervised classification requires the analyst to select training areas where he/she knows what is on the ground and then digitize a polygon within that area…
Mean Spectral Signatures
Known Conifer Area
Known Water Areac
Known Deciduous Area
Conifer
Deciduous
Water
Supervised Classification
Multispectral ImageInformation
(Classified Image)
Mean Spectral Signatures
Spectral Signature of Next Pixel to be Classified
Conifer
Deciduous
Water Unknown
Unsupervised Classification
Unsupervised classification reverses the supervised classification process. Spectral classes are grouped first, based only on the numerical information in the data and are then matched by the analyst to information classes.
Programs called Clustering algorithms are used to determine the natural groupings or structures in the data. Usually, the analyst specify how many groups or clusters are to be looked for in the data.
In addition to specifying the desired number of classes, the analyst may also specify the parameters related to separation distance among the clusters and variation with each cluster
However, algorithm for this classification operates in a two- pass mode. In the first pass, the algorithm sequentially builds class clusters. In second pass, a minimum distance classifier is applied to the whole data set on a pixel-by-pixel basis, where each pixel is assigned to one of the mean vectors created in pass 1 mode.
Unsupervised Classification
Digital Image
The analyst requests the computer to examine the image and extract a number of spectrally distinct clusters… Spectrally Distinct Clusters
Cluster 3
Cluster 5
Cluster 1
Cluster 6
Cluster 2
Cluster 4
Saved Clusters
Cluster 3
Cluster 5
Cluster 1
Cluster 6
Cluster 2
Cluster 4
Unsupervised Classification
Output Classified Image
Unknown
Next Pixel to be Classified
Unsupervised Classification
??? Water
??? Water
??? Conifer
??? Conifer
??? Hardwood
??? Hardwood
The result of the unsupervised classification is not yet information until…
The analyst determines the ground cover for each of the clusters…
Remote Sensing Applications
Forestry & Ecosystem1. Forest cover & density mapping
2. Deforestation mapping
3. Forest fire mapping
4. Wetland mapping and monitoring
5. Biomass estimation
6. Species inventoryAgriculture1. Crop type classification
2. Crop condition assessment
3. Crop yield estimation
4. Mapping of soil characteristic
5. Soil moisture estimation
Land Use/Land Cover mapping1. Natural resource management
2. Wildlife protection
3. Encroachment
Urban Planning1. Land parcel mapping
2. Infrastructure mapping
3. Land use change detection
4. Future urban expansion planning
Ocean applications1. Storm forecasting
2. Water quality monitoring
3. Aquaculture inventory and monitoring
4. Navigation routing
5. Coastal vegetation mapping
6. Oil spill
Hydrology1. Watershed mapping & management
2. Flood delineation and mapping
3. Ground water targeting
Remote Sensing Applications……cont.
Geology1. Lithological mapping
2. Mineral exploration
3. Environmental geology
4. Sedimentation mapping and monitoring
5. Geo-hazard mapping
6. Glacier mapping
Other Applications1. Flood Plain Mapping
2. Disaster Management
3. District level Planning
Land Use And Land Cover Mapping
LISS III PAN
Initial Statistics
Contrast Enhancement
Registration
Supervised Classification
Classified Image
Accuracy Assessment
Ground Data
A study on land use and land cover for a part of
Hraidwar district was carried out for the area lying
between 78007’13” E and 78016’14” E longitude and
300 N and 30008’53” N latitudes covering an area of
nearly 260 km2.
IRS-1C LISS III of April 3, 2000 was used along with
PAN image of the same date. The methodology
adopted is shown in figure.
On the basis of field visit, 11 cities
were identified. These classes are:
i). Thin forest ii). Medium forest
iii). Dense forest iv). Fallow land
v). Shrubs vi). Open land
vii). Shallow water viii). Wet land
ix). Dry sand x). Built-up-area
xi). Deep water