RS Semi Final PDF

Geometric correction is the process of IMAGE ENHANCEMENT correcting these distortions and assigning the properties (and practical value) of a Colour perception takes place in the human map to an image. eye and the associated part of the brai...

Geometric correction is the process of IMAGE ENHANCEMENT correcting these distortions and assigning the properties (and practical value) of a Colour perception takes place in the human map to an image. eye and the associated part of the brain. - Colour perception concerns our Geometric correction is required if: ability to identify and distinguish images have to be compared to either field colours, which, in turn, enables us to data or maps, identify and distinguish entities in the image data have to be compared with other real world. spatial data (e.g. in a GIS), area or distance estimates are required from Three-dimensional spaces to define colours: the image data, 1. Red Green Blue(RGB) space, based on two images from different dates are to be the additive principle of colours. compared pixel by pixel (e.g. for change2.Intensity Hue Saturation (IHS) space , analysis). which is most related to our intuitive perception of colour. 1. Panoramic distortion: the image is a 3. Yellow Magenta Cyan (YMC) space two-dimensional projection based on the subtractive principle of colours 2. Orientation: satellite images are unlikely RGB - colours is directly related to the way in to be oriented towards true north. The which computer and television monitors orientation of aerial images is determined by function. the heading of the aircraft and hence the order in which image lines were scanned or The IHS system,which refers to intensity,hue photographs taken. and saturation,more naturally reflects our sensation of colour. Intensity describes 3. Earth’s rotation: Landsat crosses the whether a colour is light or dark. equator with a heading of about 188º and the - Hue refers to the names that we give earth is rotating eastwards underneath it. As to colours: red, green, yellow, orange, a result the start of each scan line is slightly purple, etc. to the west of its predecessor. The overall - Saturation describes a colour in effect is to skew the image. terms of pale versus vivid. 4. Instrument error: this is caused by The digital image enhancement may be distortions in the optical system, non-linearity defined as some mathematical operations of the scanning mechanism and non-uniform that are to be applied to digital remote sampling rates. sensing input data to improve the visual appearance of an image for better 5. Platform instability: the altitude and interpretability or subsequent digital analysis. attitude of satellites are subject to variation, albeit quite small. By contrast, the roll, pitch Contrast enhancement involves changing and yaw of an aircraft can severely distort the original values so that more of the aerial images, and are greatly influenced by available range is used, thereby increasing pilot skill and weather conditions. the contrast between targets and their backgrounds. A map projection is the means by which the - The key to understanding contrast three dimensional surface of the Earth is enhancements is to understand the drawn on a two dimensional surface (piece of concept of an image histogram. paper or computer screen).  In raw imagery, the useful data often Ideally, a distortion-free map would have populates only a small portion of the available four major properties: range of digital values (commonly 8 bits or 256 levels). 1. Conformality: the characteristic of true shape. In other words, the shape of any small histogram is a graphical representation of area is not distorted by the projection the brightness values that comprise an image. 2. Equivalence: the characteristic of equal area. In other words, areas on one portion of Linear contrast stretch - The simplest type the map are in scale with areas in any other of enhancement portion - involves identifying lower and upper bounds from the histogram (usually 3. Equidistance: the characteristic of true the minimum and maximum distance measuring. In other words, the scale brightness values in the image) and of distance is constant over the entire map applying a transformation to stretch this range to fill the full range. 4. True direction: this means that a line - enhances the contrast in the image connecting two points on the map crosses with light toned areas appearing reference lines (e.g. meridians) at a constant lighter and dark areas appearing angle. darker 1.Geographical (spherical) coordinates are Spatial filtering encompasses another set of based on a network of lines of latitude (called digital processing functions which are used to parallels) and longitude (called meridians). enhance the appearance of an image. 2. Planar (Cartesian) coordinates are Spatial filters are designed to highlight or defined by a column and row position on a suppress specific features in an image based planar grid. The origin of a planar system is on their spatial frequency. typically located south and west of the origin of the projection. Coordinates increase going Low pass filters - it is inferred that the raw north and east. The origin of the projection is data has a random noise, whereas the other defined by coordinates referred to as the lines can provide some information with false easting and false northing. UTM is a greater clarity. planar projection. - There are a number of low pass filters, such as, mean filter, median filter, and adaptive filters geodetic datum - is a smooth mathematical surface that closely fits the mean sea-level IMAGE INTERPRETATION surface throughout an area of interest, within the area of use of a spheroid. Classification is the assignment of objects, - By qualifying the map coordinates to a features, or areas to classes based on their particular datum some compensation appearance on the imagery. can be made for irregularities in the Earth’s surface. Enumeration is the task of listing or counting discrete items visible on an image. The process of georeferencing involves two steps: Measurement, or mensuration, is an 1. selection of the appropriate type of important function in many image transformation, and interpretation problems. Two kinds of 2. determination of the transformation measurement are important. parameters. Delineation Finally, the interpreter must often The errors that remain after the delineate, or outline, regions as they are transformation are called residual errors observed on remotely sensed images. GEOMETRIC CORRECTION INDICATIVE KEYS FOR PHOTO INTERPRETER Each Sensor-platform combination is likely to have its own type of geometric image Shape is simply the geometric outline of an distortion. object. Three Common Types: 1. The Effect Of Earth Rotation Size may refer to the area of an object or to a 2. The Effect Of Earth Rotation single dimension such as the length of a road 3. Relief Displacement or airport runway. In summary, several factors must be Tone is the brightness level in a monochrome weighed against each other, depending on image or the combination of brightnesses the uses for your geometrically corrected (i.e., color), in a color image. image: Type of map: habitat map, navigational Texture describes the structure of the chart? variation in brightness within an object. Special properties to be preserved: area, distance, shape? Patterns are shapes with identifiable Type of data to be mapped: road geometric or periodic attributes. distances, ground cover? Map accuracy: what is acceptable? Shadows are often thought of in a negative Scale: how big an area? sense by interpreters be- cause it can be difficult to see objects in deep shadow. Relief Displacement shows up specifically in camera images of large scale if there is Site refers to the geographic location of a significant terrain relief or if there are high target or the location of one feature relative to protruding objects. another. Distortion arises from scanner Association specifies the occurrence of characteristics and their interaction with the certain objects or features, usually without the airborne platform or satellite orbital geometry strict spatial arrangement implied by pattern. and figure of the earth Stereoscopy is the ability to derive distance There are three components to the information (or in the case of aerial process: photography, height information) from two 1. Selection of suitable mathematical images of the same scene. distortion model(s) 2. Coordinatetransformation Stereoscopes are devices that facilitate 3. Resampling(interpolation) stereoscopic viewing of aerial photographs. -This simplest and most common is the These three components are collectively pocket stereoscope known as warping Scale is an expression of the relationship of Various terms are used to describe the image distance between two points and geometric correction of imagery, and it is the actual distance between the two worthwhile defining them before corresponding points on the ground. proceeding: 1. Registration - The alignment of one image A second method of specifying scale is the to another image of the same area. bar scale,which simply labels a line with subdivisions that show ground distances. 2. Rectification - The alignment of an image to a map so that the image is planimetric, just the representative fraction (RF),is more like the map. Also known as georeferencing. widely used and often forms the preferred An example rectification of Landsat TM method of reporting image data to map coordinates - ratio between image distance and ground distance 3. Geocoding - special case of rectification that includes scaling to a uniform, standard pixel GSI. The use of standard pixel sizes FINALS and coordinates permits convenient “layering” of images from different sensors İNFORMATİON EXTRACTION and maps in a GIS. Information extraction based on visual 4. Orthorectification - Correction of the image interpretation. image, pixel-by-pixel, for topographic - Typical examples of this approach are distortion. The result, in effect, is that every visual interpretation methods for land pixel appears to view the earth from directly use or soil mapping. above, i.e., the image is in an orthographic - Also acquiring data for topographic projection mapping from aerial photographs is still largely based on visual quadratic polynomial ( N equals 2) is interpretation. sufficient for most problems in satellite remote sensing, where the terrain relief is small and Information extraction based on the FOV is not large. semi-automatic processing by the computer. GCPs should have the following - Examples include automatic characteristics: generation of DTMs, digital image classification and calculation of high contrast in all images of interest surface parameters. small feature size unchanging over time all are at the same elevation (unless Accuracy defines "correctness"; it measures topographic relief is being specifically the agreement between a standard assumed addressed) to be correct and a classified image of unknown quality. - An example of GCP selection and polynomial fitting can be generated Precision defines a measure of the from the aerial photograph and map. sharpness (or certainty) of a measurement - A set of six GCPs and four test GPs are shown on the photograph and the Classification error is the assignment of a scanned map pixel belonging to one category (as determined by ground observation) to another Since the (x,y) coordinates will not be integer category during the classification process. values in general, a new pixel must be estimated between existing pixels by an Site-Specific Accuracy, is based on the interpolation process known as resampling detailed assessment of agreement between the map and reference data at specific Maps are the spatial framework for analysis locations of remote sensing images. The standard form for reporting site-specific Resampling can be thought of as error is the error matrix, sometimes referred convolution of the distorted image with a to as the confusion matrix because it moving window function, as in spatial filtering. identifies not only overall errors for each category but also misclassifications (due to The fastest scheme for calculating a confusion between categories) by category. resampled pixel is a nearest-neighbor assignment(some called zero-order MEASUREMENT OF MAP interpolation) ACCURACY(COMPILING THE ERROR MATRIX) - is very important that the validation The Wiener Filter minimizes the error in samples and the map be well estimating the ideal image from the noisy coregistered to one another. image by a linear filter operation. Examination of the error matrix reveals, for Local Noise refers to individual bad pixels each category, and lines, typically caused by data - errors of omission transmission loss, sudden detector saturation - errors of commission. or other intermittent electronic problems One of the most widely used measures of Global periodic noise (commonly called accuracy is the overall accuracy, a report of coherent noise) is a spurious, repetitive the overall proportion of correctly classified pattern that has consistent characteristics pixels in the image or in the sample used to throughout an image. construct the matrix. Correction for detector striping, called Hay (1979) stated that it is necessary to destriping consider five questions to thoroughly under- stand the accuracy of a classification: Geocoding is required when different images 1. What proportion of the classification need to be combined or when the image data decision is correct? is used in a GIS environment that requires all 2. What proportion of assignments to a given data to be stored in the same map projection. category is correct? - Geocoding is georeferencing with 3. What proportion of a given category is subsequent resampling of the image correctly classified? raster. 4. Is a given category overestimated or underestimated? RMSE ( Root mean square error ) is a 5. Are errors randomly distributed? measure of the magnitude of error but it does not incorporate bias since the squaring The error matrix is an example of a more eliminates the direction of the error. general class of matrices, known as - commonly used to document vertical contingency tables accuracy for digital elevation models (DEMs). K (kappa) is a measure of the difference - Root mean square error (RMSE) between the observed agreement between measures the error between a two maps mapped point and a measured ground THERMAL REMOTE SENSING position. Thermal Remote Sensing (TRS) is based on the measuring of electromagnetic radiation (EM) in the infrared region of the spectrum. Most commonly used are the intervals from 3–5 μm (MIR) and 8–14 μm (TIR) Planck’s Radiation Law describes the amount of emitted energy per wavelength depending on the objects temperature Blackbody The three laws described above are, strictly speaking, only valid for an ideal radiator, which we refer to as a blackbody (BB). - perfect absorber and a perfect radiator in all wavelengths. Greybody Materials that absorb and radiate only a certain fraction compared to a blackbody are called greybodies. - fraction is a constant for all wavelengths. Hence, a greybody curve is identical in shape to a blackbody curve Selective Radiator A third group are the selective radiators. They also radiate only a certain fraction of a blackbody, but this fraction changes with wavelength. - radiate perfectly in some wavelengths, Emissivity The fraction of energy that is radiated by a material compared to a true blackbody is also referred to as emissivity (ελ). Radiant temperature The actual measurements acquired by a TIR sensor will be in units of spectral radiance (W m−2 sr−1 μm−1) that reaches the sensor for a certain wavelength band. Kinetic temperature The radiant temperature calculated from the radiant energy emitted is in most cases smaller than the true, kinetic temperature (Tkin) that we could measure with a contact thermometer Types of Detectors and Their Sensitivity Indium antimonide (InSb): Sensitive around 5 μm (mid-infrared). Mercury-doped germanium (Ge:Hg): Sensitive around 10 μm (far-infrared). Mercury cadmium telluride (MCT): Sensitive across 8-14 μm. Detector Sensitivity and Resolution High sensitivity = captures fine details. Low sensitivity = only shows large temperature differences. Higher detail (spatial resolution) can reduce brightness accuracy (radiometric resolution) due to weaker signals. A radiometer is a sensor that measures the intensity of radiation received within a specified wavelength A device known as a chopper is capable of interrupting the radiation that reaches the detector. - The chopper consists of a slotted disk or similar device rotated by an electrical motor Spatial Resolution Of A Radiometer Is determined by an instantaneous field of view (IFOV), which is in turn controlled by the sensor’s optical system, the detector, and flying altitude Microwave radiometers are very sensitive instruments tailored to receive and record radiation in the range from about 0.1 mm to 3 cm. - consists of a sensitive receiving instrument typically in the form of a horn- or dish-shaped antenna Thermal scanners sense radiances of features beneath the aircraft flight path and produce digital and/or pictorial images of that terrain. TYPES OF THERMAL SCANNER Object-plane scanners view the landscape by means of a moving mirror that oscillates at right angles Image-plane scanners use a wider field of view to collect a more comprehensive image of the landscape infrared scanning system consists of a scanning unit, with a gyroscopic roll connection unit, infrared detectors