«Текущий контроль результатов изучения дисциплины Примеры вопросов и заданий при защите лабораторных работ Для подготовки заданий имеется архив из 52 статей на ...»
Текущий контроль результатов изучения дисциплины
Примеры вопросов и заданий при защите лабораторных работ
Для подготовки заданий имеется архив из 52 статей на английском языке.
Что обозначает временное, пространственное, спектральное и радиометрическое разрешение космических снимков?
Опишите основные характеристики данных со спутника Landsat.
Опишите систему регистрации снимков Landsat WRS2.
Какая информация хранится в спецификации названий файлов, предоставляемых USGS?
В чем состоит отличие данных, хранящихся в разных файлах мультиспектрального снимка?
Расшифруйте название файла. Пример: Entity ID: LT51560192011294KHC00
Опишите представленный ниже интерфейс выбора снимков.
2190755334000Какие основные критерии нужно задавать при поиске и заказе снимков на сайте http://earthexplorer.usgs.gov/Какие дополнительные критерии можно задавать при поиске и заказе снимков на сайте http://earthexplorer.usgs.gov/Опишите, в чем различие данных, находящихся в наборах данных
L7 ETM+ SLC-off и L7 ETM+ SLC-on?
Оценить, сколько снимков Landsat 5 за период с 01.06.2007 по 30.08.2007 с облачным покрытием 0 % с Path=155 и Row=19 и уровнем обработки L1 предоставляет сервис earthexplorer.usgs.gov
Задание 1. Перевод отрывка текста по теме лабораторной работы.
L5 was developed by the National Aeronautics and Space Administration (NASA) and launched in March 1984.
In general terms, an image is a digital picture or representation of an object. Remotely sensed image data are digital representations of the Earth. Image data are stored in data files, also called image files, on magnetic tapes, computer disks, or other media. The data consist only of numbers. These representations form images when they are displayed on a screen or are output to hardcopy. Each number in an image file is a data file value. Data file values are sometimes referred to as pixels. The term pixel is abbreviated from picture element. A pixel is the smallest part of a picture (the area being scanned) with a single value. The data file value is the measured brightness value of the pixel at a specific wavelength.
Raster image data are laid out in a grid similar to the squares on a checkerboard. Each cell of the grid is represented by a pixel, also known as a grid cell. In remotely sensed image data, each pixel represents an area of the Earth at a specific location. The data file value assigned to that pixel is the record of reflected radiation or emitted heat from the Earth’s surface at that location.
Resolution is a broad term commonly used to describe:
• the number of pixels you can display on a display device, or
• the area on the ground that a pixel represents in an image file.
These broad definitions are inadequate when describing remotely sensed data. Four distinct types of resolution must be considered:
• spectral—the specific wavelength intervals that a sensor can record
• spatial—the area on the ground represented by each pixel
• radiometric—the number of possible data file values in each band (indicated by the number of bits into which the recorded energy is divided)
• temporal—how often a sensor obtains imagery of a particular area
These four domains contain separate information that can be extracted from the raw data.
ЛАБОРАТОРНАЯ РАБОТА 2.
В чем состоит преимущество использования мультиспектральных снимков?
Какие типы синтезирования цвета применяются для лучшей визуальной интерпретации растительности по КС и почему?
Опишите полученные результаты слияния мультиспектральных данных с панхроматическими данными.
Задание. Перевод отрывка текста по теме лабораторной работы.
Landsat TM sensors have seven bands with a spatial resolution of 28.5 m. SPOT panchromatic has one broad band with very good spatial resolution—10 m.
Combining these two images to yield a seven-band data set with 10 m resolution provides the best characteristics of both sensors.
A number of models have been suggested to achieve this image merge. Welch and Ehlers (1987) used forward-reverse RGB to IHS transforms, replacing I (from transformed TM data) with the SPOT panchromatic image. However, this technique is limited to three bands (R,G,B).
Spectral calibrations between two sensors
The use of finer spatial resolution data is an essential step to integrate ground measurements with coarse spatial resolution data. To perform this step, the relationship of spectral responses was examined. We conducted internal spectral calibrations between 30-m ETM+ and 1-km MODIS data due to the differences in spatial and spectral resolutions between the two sensors. We examined the correlations of blue, red, near infrared, middle infrared reflectance after finer ETM+ data were aggregated to the same resolution (1-km) of MODIS data within the area of ETM+ scene. The resulting spectral relationships for the identified bands within the ETM+ scene were used for calibrating MODIS data across the region. These bands are either being used directly as driving variables in the model or indirectly for calculating various vegetation indices that drive the model. Such
calibrations are necessary because:
1) the models are established using ETM+ data, and
2) remotely sensed information between the 2 sensors is not identical, even for the same targets.
ЛАБОРАТОРНАЯ РАБОТА 3.
Обоснуйте необходимость геометрической коррекции КС.
Как следует выбирать степень полинома при геометрической коррекции КС?
Приведите требования в расстановке точек при геометрической коррекции изображений.
Обоснуйте необходимый минимум наземных контрольных точек для коррекции изображений с помощью полиномов 2 степени.
Прокомментируйте точность выполненной привязки с использованием инструмента географического связывания.
Прокомментируйте точность выполненной привязки с использованием инструмента «Прозрачность».
Прокомментируйте точность расстановки точек путем замены типа точки.
Задание. Перевод отрывка текста по теме лабораторной работы.
Raw, remotely sensed image data gathered by a satellite or aircraft are representations of the irregular surface of the Earth. Even images of seemingly flat areas are distorted by both the curvature of the Earth and the sensor being used. This chapter covers the processes of geometrically correcting an image so that it can be represented on a planar surface, conform to other images, and have the integrity of a map. A map projection system is any system designed to represent the surface of a sphere or spheroid (such as the Earth) on a plane. There are a number of different map projection methods. Since flattening a sphere to a plane causes distortions to the surface, each map projection system compromises accuracy between certain properties, such as conservation of distance, angle, or area. For example, in equal area map projections, a circle of a specified diameter drawn at any location on the map represents the same total area. This is useful for comparing land use area, density, and many other applications. However, to maintain equal area, the shapes, angles, and scale in parts of the map may be distorted (Jensen 1996). There are a number of map coordinate systems for determining location on an image. These coordinate systems conform to a grid, and are expressed as X,Y (column, row) pairs of numbers. Each map projection system is associated with a map coordinate system.
Rectification is the process of transforming the data from one grid system into another grid system using a geometric transformation. While polynomial transformation and triangle-based methods are described in this chapter, discussion about various rectification techniques can be found in Yang (1997). Since the pixels of the new grid may not align with the pixels of the original grid, the pixels must be resampled.
Resampling is the process of extrapolating data values for the pixels on the new grid from the values of the source pixels. The pixels must be resampled to conform to the new grid.
Georeferencing refers to the process of assigning map coordinates to image data. The image data may already be projected onto the desired plane, but not yet referenced to the proper coordinate system. Rectification, by definition, involves georeferencing, since all map projection systems are associated with map coordinates. Image-to-image registration involves georeferencing only if the reference image is already georeferenced. Georeferencing, by itself, involves changing only the map coordinate information in the image file. The grid of the image does not change. Geocoded data are images that have been rectified to a particular map projection and pixel size, and usually have had radiometric corrections applied. It is possible to purchase image data that is already geocoded.
Disk rectification involves:
• rearranging the pixels of the image onto a new grid, which conforms to a plane in
the new map projection and coordinate system
• inserting new information to the header of the file, such as the upper left corner
map coordinates and the area represented by each pixel
Usually, rectification is the conversion of data file coordinates to some other grid and coordinate system, called a reference system. Rectifying or registering image data on disk involves the following general steps, regardless of the application:
1. Locate GCPs.
2. Compute and test a transformation.
3. Create an output image file with the new coordinate information in the header.
Images can be rectified on the display (in a Viewer) or on the disk. Display rectification is temporary, but disk rectification is permanent, because a new file is created.
ЛАБОРАТОРНАЯ РАБОТА 4.
Для каких целей обычно применяют неконтролируемую классификацию аэрокосмических снимков?
Опишите принципы работы алгоритма ISODATA.
В чем заключается основная особенность алгоритмов неконтролируемой классификации в сравнении с контролируемой?
Постройте легенду тематической карты классификации породного состава лесов на основании представленной ниже карты лесной таксации и ее легенды. Обоснуйте выбор классов для легенды.
В каком случае полученные классы следует объединить?
Задание. Перевод отрывка текста по теме лабораторной работы.
Unsupervised training is more computer-automated. It enables you to specify some parameters that the computer uses to uncover statistical patterns that are inherent in the data. These patterns do not necessarily correspond to directly meaningful characteristics of the scene, such as contiguous, easily recognized areas of a particular soil type or land use. They are simply clusters of pixels with similar spectral characteristics. In some cases, it may be more important to identify groups of pixels with similar spectral characteristics than it is to sort pixels into recognizable categories.
Unsupervised training is dependent upon the data itself for the definition of classes. This method is usually used when less is known about the data before classification. It is then the analyst’s responsibility, after classification, to attach meaning to the resulting classes (Jensen 1996). Unsupervised classification is useful only if the classes can be appropriately interpreted.
ISODATA Clustering Parameters
To perform ISODATA clustering, you specify:
• N - the maximum number of clusters to be considered. Since each cluster is the basis for a class, this number becomes the maximum number of classes to be formed. The ISODATA process begins by determining N arbitrary cluster means.
Some clusters with too few pixels can be eliminated, leaving less than N clusters.
• T - a convergence threshold, which is the maximum percentage of pixels whose class values are allowed to be unchanged between iterations.
• M - the maximum number of iterations to be performed.
Initial Cluster Means
On the first iteration of the ISODATA algorithm, the means of N clusters can be arbitrarily determined. After each iteration, a new mean for each cluster is calculated, based on the actual spectral locations of the pixels in the cluster, instead of the initial arbitrary calculation. Then, these new means are used for defining clusters in the next iteration. The process continues until there is little change between iterations (Swain 1973).
The initial cluster means are distributed in feature space along a vector that runs
between the point at spectral coordinates (1-1, 2-2, 3-3,... n-n) and the
coordinates (1+1, 2+2, 3+3,... n+n). The initial cluster means are evenly distributed between (A-A, B-B) and (A+A, B+B).
ЛАБОРАТОРНАЯ РАБОТА 5.
Задания и разговорная часть
Оцените точность классификации по полученной матрице ошибок классификации.
Оцените ошибки пропуска и добавления.
Рассчитайте индекс Каппа.
Оцените разделимость обучающих выборок по трансформируемой дивергенции.
Какие каналы наиболее подходят для классификации КС на основании значений трансформируемой дивергенции.
Оцените разделимость обучающих выборок по гистограммам в каждом канале.
Какие каналы наиболее подходят для классификации КС на основании построенных гистограмм?
Задание 4. Перевод отрывка текста по теме лабораторной работы.
Supervised training requires a priori (already known) information about the data, such as:
• What type of classes need to be extracted? Soil type? Land use? Vegetation?
• What classes are most likely to be present in the data? That is, which types of land cover, soil, or vegetation (or whatever) are represented by the data?
In supervised training, you rely on your own pattern recognition skills and a priori knowledge of the data to help the system determine the statistical criteria (signatures) for data classification. To select reliable samples, you should know some information – either spatial or spectral – about the pixels that you want to classify.
The location of a specific characteristic, such as a land cover type, may be known through ground truthing. Ground truthing refers to the acquisition of knowledge about the study area from field work, analysis of aerial photography, personal experience, etc. Ground truth data are considered to be the most accurate (true) data available about the area of study. They should be collected at the same time as the remotely sensed data, so that the data correspond as much as possible. However, some ground data may not be very accurate due to a number of errors and inaccuracies.
A process is iterative when it repeats an action. The objective of the ERDAS IMAGINE system is to enable you to iteratively create and refine signatures and classified image files to arrive at a desired final classification. The ERDAS IMAGINE classification utilities are tools to be used as needed, not a numbered list of steps that must always be followed in order.
The total classification can be achieved with either the supervised or unsupervised
methods, or a combination of both. Some examples are below:
• Signatures created from both supervised and unsupervised training can be merged and appended together.
• Signature evaluation tools can be used to indicate which signatures are spectrally similar. This helps to determine which signatures should be merged or deleted. These tools also help define optimum band combinations for classification. Using the optimum band combination may reduce the time required to run a classification process.
• Since classifications (supervised or unsupervised) can be based on a particular area of interest (either defined in a raster layer or an.aoi layer), signatures and classifications can be generated from previous classification results.
Typically, the accuracy of a crisp land cover classification from remote sensing
images is assessed using error matrix, which is defined as cross-tabulation of pixels of agreement and disagreement arranged in a two dimensional matrix of rows and columns. The rows represent class allocation of pixels in classified image and columns represent their corresponding class allocation as inferred from ground or reference data albeit both are interchangeable. As the error matrix represents a sample of testing pixels, its creation is highly dependent on a sound sampling design procedure. The elements of error matrix are used to derive a number of accuracy measures such as overall, user’s and producer’s accuracy, kappa coefficient etc. The information contained in error matrix can also be used for estimation of areas of land cover classes using any of the five techniques implemented here.
Traditional spectral classification of remotely sensed images applied on a pixel-by-pixel basis ignores the potentially useful spatial information between the values of proximate pixels. For some 30 years the spatial information inherent in remotely sensed images has been employed, albeit by a limited number of researchers, to enhance spectral classifcation. This has been achieved primarily by filtering the original imagery to (i) derive texture `wavebands' for subsequent use in classifcation or (ii) smooth the imagery prior to (or after) classifcation. Recently, the variogram has been used to represent formally the spatial dependence in remotely sensed images and used in texture classifcation in place of simple variance filters. However, the variogram has also been employed in soil survey as a smoothing function for unsupervised classifcation. In this review paper, various methods of incorporating spatial information into the classifcation of remotely sensed images are considered. The focus of the paper is on the variogram in classifcation both as a measure of texture and as a guide to choice of smoothing function. In the latter case, the paper focuses on the technique developed for soil survey and considers the modifcation that would be necessary for the remote sensing case.
One of the simplest approaches to supervised classification is the minimum-distance-to-mean classifer, also referred to as the k-means or (here) c-means classifer. This classifer utilises the Euclidean distances in (spectral) feature space between (i) the pixels to be classifed and (ii) the class means (obtained from training data).
In the univariate case, a training site is selected which belongs to a desired class and the values for all pixels at the site are averaged to obtain the class mean. This procedure is repeated for all other classes of interest. Each pixel in the remainder of the image may then be allocated to the class mean to which it is nearest in univariate feature space. In a simple empirical approach presented by Carr (1996) this allocation is executed through the prior construction of a simple look-up table such that any pixel value will have associated with it the destination class without the need to compare the distances to all class means. This procedure is readily extended to multivariate feature space where the distances to class means may be obtained by Euclidean geometry.