Digital Commons Network^™

A Land Use Classification System For Use With Remote-Sensor Data, Ernest E. Hardy, James R. Anderson

LARS Symposia

New demands on our land resources require more stringent controls and management practices. The administration of these controls requires better and more frequent information concerning land use. Although new tools became available to aid in acquiring and processing the data, a major lack in uniform techniques for identifying the land use was a major problem. Creation of a more standard form of classification of land use, based on the capabilities inherent in the various forms of remote sensors and other data sources was a necessary step.

A classification has been created, and published in preliminary form, by the Geological Survey …

Combining Human And Computer Interpretation Capabilities To Analyze Erts Imagery, J. D. Nichols, W. M. Senkus

LARS Symposia

The human has the ability to quickly delineate gross differences in land classes, such as wildland, urban and agriculture on appropriate ERTS images, and to further break these gross classes into meaningful subclasses. I; agricultural areas, the subclasses can be delineated on the basis of general tone and texture differences that relate to crop type and field size. In the wildland areas, delineations can also be made, based on tone and texture, which represent general vegetation systems, such as grasslands, brush, trees, and barren areas. The computer, however, can more efficiently analyze point-by-point spectral information and localized textural information which …

The Role Of Computer Networks In Remote Sensing Data Analysis, P. H. Swain, T. L. Phillips, J. C. Lindenlaub

LARS Symposia

It has been hypothesized that computer networks can be used to make data processing facilities available to the remote sensing community both quickly and effectively. An experiment to test this hypothesis is being conducted by the Laboratory for Applications of Remote Sensing at Purdue University, with the participation of potential users at several remote sites. Initial indications have been highly favorable, although final evaluation awaits further experience and the accumulation of usage data.

Erts-1 Aircraft Support, 24-Channel Mss Cct Experiences And Land Use Classification Results, A. J. Richardson, M. R. Gautreaux, C. L. Wiegand

LARS Symposia

Aircraft multispectral scanner (MSS) data collected by the NASA 24-channel MSS on July 26, 1972 (Mission 207) over the USDA, Research Farm at Weslaco, Texas, were evaluated for quality and for crop, soil, and water discrimination. The standard deviations for each of the 24 channels for a uniform surface, a water reservoir, were used as an indicator of system noise. By this criterion, channels 22, 20, 15, qnd 21 were of low quality. Based on the ratio of odd to even numbers in all channels, the conclusion was reached that the data are 7-bit precision. An optimum channel selection program …

Urban Land-Use Mapping By Machine Processing Of Erts-1 Multispectral Data: A San Francisco Bay Area Example, Richard Ellefsen, Philip H. Swain, James R. Wray

LARS Symposia

Results are promising of an experiment to map land use in an urban area by automatic digital processing of ERTS-1 data. Computer-maps of a large segment of the San Francisco-Oakland and San Jose Urbanized Areas have been produced at a scale of 1:24,000 using a segment of an ERTS-l frame reformatted to correct skewness and scale. An area of some 6,500 square kilometers was also mapped at 1:48,000 (a one-fourth sample). For both scales, urban areas were separated from rural -- using a photo interpretation procedure -- to solve problems of the spectral similarity of functionally different land uses and …

Analysis Of Remotely Sensed Data For Detecting Soil Limitations, L. A. Benson, C. J. Frazee, F. A. Waltz

LARS Symposia

During 1971 and 1972 a detailed study was conducted on a fallow field in the proposed Oahe Irrigation Project to determine the relationship between the tonal variation observed on aerial photographs and the properties of eroded soil. Correlation and regression analysis of digitized, multiemulsion, color infrared film (2443) data and detailed field data revealed a highly significant correlation between film transmittance and several soil properties indicative of the erosion limitation. Computer classification of the multiemulsion film data resulted in maps portraying the eroded soil and the normal soil. Both correlation and computer classification results were best using the reflectance data …

The Use Of The Modified Cholesky Decomposition In Divergence And Classification Calculations, D. L. Van Rooy, M. S. Lynn, C. H. Snyder

LARS Symposia

This report analyzes the use of the modified Cholesky decomposition technique as applied to the feature selection and classification algorithms used in the analysis of remote sensing data (e. g.•as in LARSYS). This technique is approximately 30% faster in classification and a factor of 2- 3 faster in divergence, as compared with LARSYS. Also numerical stability and accuracy are slightly improved. Other methods necessary to deal with numerical stability problems are briefly discussed.

Deriving Spectral And Spatial Features To Establish A Hierarchical Classification System, James E. Skaley, Randolph J. Hoffman

LARS Symposia

Automatic processing of remotely sensed data has to date been constrained to using training sets to classify a small number of categories within the context of a limited geographical area.

In order to promote a more flexible user-oriented data processing system, a hierarchical taxonomic structure is proposed. This structure incorporates data inputs from several different sensors together with a priori information on the characteristics of different materials of interest to facilitate efficient design of feature sets to classify those materials. A Boolean approach may be used to assign these feature sets including both spectral and spatial criteria to different hierarchical …

Feature Extraction Of Multispectral Data, R. B. Crane, T. Crimmins, J. F. Reyer

LARS Symposia

A method is presented for feature extraction of multispectral scanner data. Non-training data is used to demonstrate the reduction in processing time that can be obtained by using feature extraction rather than feature selection.

The Jsc Clustering Program Isocls And Its Applications, E. P. Kan, W. A. Holley, H. D. Parker Jr.

LARS Symposia

The clustering program ISOCLS developed at the Johnson Space Center, Houston, Texas, has been extensively used in the pattern analysis and classification of remote sensor data collected by aircraft and by the Earth Resources Technology Satellite ERTS-1. This paper discusses the theory behind this clustering algorithm. Several new ideas that have been incorporated in ISOCLS are discussed.· Among these are the novel philosophy of operation behind the procedure, which assumes that a population (i.e. a class or a cluster) can be treated as the union of an appropriate number of subpopulations, and the termination of the clustering program by a …

Mapping Soil Associations Using Erts Mss Data, Jan E. Cipra

LARS Symposia

Multispectral imagery from ERTS was compared to a soil association map-of Tippecanoe County, Indiana, based on a conventional field survey. Three methods of processing single and multiple wavelength bands were applied to imagery collected on three dates. Correction of geometric distortions in computer-processed imagery was investigated. Four to six groupings of soil associations could be delineated on ERTS imagery. Some associations were more readily delineated using visible wavelengths, some more readily delineated using infrared wavelengths. In general, MSS band 7 was found to give more soils information than MSS band s. Imagery of single wavelength bands displayed on a television …

A Method For Digital Image Registration Using A Mathematical Programming Technique, Stanton S. Yao

LARS Symposia

A new algorithm based on a nonlinear programming technique to correct the geometrical distortions of one digital image with respect to another is discussed. This algorithm promises to be superior to existing ones in that it is capable of treating , localized differential scaling, translational and rotational errors over the whole image plane. A series of piece-wise "rubber-sheet" approximations are used, constrained in such a manner that a smooth approximation over the entire image can be obtained. The theoretical derivation is included. The result of using the algorithm to register four channel S065 Apollo IX digitized photography over Imperial Valley, …

An Iterative Approach To The Feature Selection Problem, Henry P. Decell, John A. Quirein

LARS Symposia

The B-average divergence for m-distinct ~asses, resulting from the linear transformation y = Bx, is proposed as a feature selection criterion, where B is a k by n matrix of rank k <= n. It is shown that if the B-average divergence resulting from B is large enough, then the probability of misclassification, considered as a function of the class of all k by n matrices, is essentially minimized by B. A computer program, utilizing a gradient procedure, is developed to numerically maximize the B-average divergence and results are presented for the Cl flight line. For this example, corresponding to 9-distinct classes, most of the discriminatory information is found to lie in a 3-dimensional subspace, defined by an appropriately chosen 3 by 12 matrix B.

Machine Boundary Finding And Sample Classification Of Remotely Sensed Agricultural Data, J. N. Gupta, R. L. Kettig, D. A. Landgrebe, P. A. Wintz

LARS Symposia

Initially, methods for analyzing earth observational data involved the use of only spectral variations measured from the scene. The work reported in this paper provides a method for making some use of spatial variations as well. The results of some preliminary tests of this new method show significant improvements in accuracy.

Computerized Interpretation Of Erts Data For Forest Management, Leonard Kirvida

LARS Symposia

Multispectral and spatial features are evaluated for automatic delineation of forest and associated land types. The principal components algorithm was used for determining the efficacy of multispectral bands in making class separations. Four spatial algorithms were evaluated for texture measurements. A thematic map was generated using four multispectral bands as features. Clusters were also used for generating a thematic map.

Multivariate Interactive Digital Analysis System (Midas): A New Fast Multispectral Recognition System, F. Kriegler, R. Marshall, S. Lampert, M. Gordon, C. Connell, R. Kistler

LARS Symposia

As a consequence of the quantity and rate of acquiring multispectral data, fast and inexpensive processing is required before practical use may be made of this technique. The speed of this machine is such that an ERTS frame could be processed in forty seconds if the proper tapes were available. Such speeds now make the man operating the system the major limiting factor in performance and requires an experimental matching of the machine to the operator. The system is a prototype, multiple-pipeline digital processor mechanizing the multivariate-Gaussian, maximum-likelihood decision algorithm operating at 2 x 10(5) pixels/second. It incorporates displays and …

Prefix Codes, Trees And Automata, Jean-Louis Lassez

Department of Computer Science Technical Reports

No abstract provided.

Techniques For Image Registration, W. F. Webber

LARS Symposia

Techniques are developed for determining spatial or geometric distortions between two images of the same scene. The first procedure is iterative linearized least squares estimation (LLSE) for determining small geometric distortions between images. Error variances for these estimators are derived which are interpreted as noise-to-signal ratios for translational and rotational registration. The natural measure of the signal strength of an image for translational registration obtained from these variances is used to establish threshold settings in a new algorithm for fast translational registration. This algorithm belongs to the class of sequential similarity detection algorithms (SSDA's) recently developed for translational registration. Finally, …

Multitemporal Geometric Distortion Correction Utilizing The Affine Transformation, R. A. Emmert, C. D. Mcgillem

LARS Symposia

In the analysis of multitemporal remotely sensed imagery, it is necessary to place these data into registration. To implement this operation the data are divided into subimages, and the misregistration between the data subsets is modeled by an affine transformation. The properties of the Fourier transform of a two-dimensional function under the affine transformation are given, and examples of these relations between the spatial and spatial frequency domains are shown. Techniques for the estimation of the coefficients of the distortion model using the spatial frequency information are developed, and an example of the use of this method for the correction …

Iterative Techniques To Estimate Signature Vectors For Mixture Processing Of Multispectral Data, Pete Salvato Jr

LARS Symposia

Mixture processing of remotely sensed multispectral scanner data involves estimating the percent coverage of individual crops or species contained within the instantaneous field of view of the scanner. In recent years, various mixture processing algorithms have been proposed to solve the so-called "mixture problem". All of the proposed algorithms require, as inputs, the spectral signatures of the various species observed. Often it is extremely difficult to obtain the required spectral signatures of individual species.

In this paper, two methods for obtaining the required spectral signatures for a particular mixture model are considered. For the model considered, the spectral signatures become …

On The Management And Processing Of Earth Resources Information, C. W. Skinner, R. C. Gonzalez

LARS Symposia

The basic concepts of a recently-completed large-scale earth resources information system plan are reported. Attention is focused throughout the paper on the information management and processing requirements. After the development of the principal system concepts, a model system for implementation at the state level is discussed.

The Design And Use Of Special Purpose Processors For The Machine Processing Of Remotely Sensed Data, G. R. Allen, L. O. Bonrud, J. J. Cosgrove, R. M. Stone

LARS Symposia

The processing of the expected volumes of remotely sensed data will overburden the available computer resources. One solution to the problem is the use of special purpose processors. This paper describes two such processors which are suitable for processing remotely sensed data. Also, examples of the use of the processors on specific problems encountered in remotely sensed data are described.

Land Use Classification Of Marion County, Indiana By Spectral Analysis Of Digitized Satellite Data, William J. Todd, Marion F. Baumgardner

LARS Symposia

Multispectral scanner data, obtained over Marion County (Indianapolis), Indiana at an altitude of 915 kilometers, were analyzed by computer-implemented techniques to evaluate the utility of satellite data for urban land use classification. Several land use classes, such as commerce/industry, single-family (newer) residential, trees, and water exhibited spectrally separable characteristics and were identified with greater than 90 per cent accuracy. Difficulties were encountered in the spectral separation of grassy (open, agricultural) areas and multi-family (older) housing. The confusion between these two classes was largely eliminated, however when spectral characteristics of samples (instead of individual data points) were considered. Another solution to …

Feature Selection Via An Upper Bound (To Any Degree Tightness) On Probability Of Misclassification, Cecil R. Hallum

LARS Symposia

Currently, many techniques exist for feature selection purposes which are related but, unfortunately, in an indeterminable way to the probability of misclassification. In this paper a procedure is presented which yields an upper bound (to any degree tightness) on the probability of misclassification in sample Gaussian maximum likelihood classification between each pair of categories in p-dimensional space. The technique permits features to be selected so that the optimal q (q <= p) features have the property that no other subset of q features yield a smaller value to the upper bound on the probability of misclassification.

A computer-assessable transformation is utilized which permits a multiple integral over the misclassification region in p-dimensional space to be approximated, to any degree of accuracy, by the product of p …

Machine Processing Of Erts And Ground Truth Data, Robert H. Rogers, Keith Peacock

LARS Symposia

Results achieved by ERTS-Atmospheric Experiment PR303, whose objective is to establish a radiometric calibration technique, are reported. This technique, which determines and removes solar and atmospheric parameters that degrade the radiometric fidelity of ERTS data, transforms the ERTS sensor radiance measurements to absolute target reflectance signatures. A Radiant Power Measuring Instrument (RPMI) and its use in determining atmospheric parameters needed for ground truth are discussed. The procedures used and results achieved in machine processing ERTS computer- compatible tapes and atmospheric parameters to' obtain target reflectance are reviewed.

Estimation Of Proportions Of Objects And Determination Of Training Sample-Size In A Remote Sensing Application, R. S. Chhikara, P. L. Odell

LARS Symposia

One of the problems in remote sensing is estimating the expected proportions of certain categories of objects which cannot be observed directly or distinctly. For example, a multi-channel scanning device may fail to observe objects because of obstructions blocking the view, or different categories of objects may make up a resolution element giving rise to a single observation. This will require ground truth on any such categories of objects for estimating their expected proportions associated with various classes represented in the remote sensing data. Considering the classes to be distributed as multivariate normal with different mean vectors and common covariance, …

Numerical Classification Procedures, Glen C. Gustafson

LARS Symposia

A number of up-to-date numerical classification techniques are described. These include the orthogonal and oblique factor analysis methods, and the unweighed pair-group cluster analysis procedure. The techniques are applied to morphometric data from 159 small drainage basins from two geographical regions. Transformation techniques to achieve the normal distribution with respect to symmetry are applied.

Classification Of Turbidity Levels In The Texas Marine Coastal Zone, E. A. Weisblatt, J. B. Zaitzeff, C. A. Reeves

LARS Symposia

.The unsupervised Iterative Self-Organizing Clustering System (ISOCLS) and the supervised Earth Resources Interactive Processing System (ERIPS) were used to detect, delineate and classify near-surface turbidity patterns in the Galveston and Trinity Bays, Texas and adjacent coastal waters. Data used in the analysis was ERTS-l Multispectral Scanner (MSS) digital data in the visible spectral bands from 0.5 to 1.1 micrometers, and related in situ water measurements.

Theoretical considerations suggest that because solar radiation attenuates with water depth and water constituents as a function of wavelength, classification of turbidity levels based on spectral characteristics is a classification based on spectral signatures from …

Information Preserving Coding For Multispectral Data, J. R. Duan, P. A. Wintz

LARS Symposia

A general formulation of the data compression system is presented. A method of instantaneous expansion of quantization levels by reserving two codewords in the codebook to perform a folding over in quantization is implemented for error free coding of data with incomplete knowledge of the probability density function. Results for simple DPCM with folding and an adaptive transform coding technique followed by a DPCM technique are compared using ERTS-l data.

Remote Wind Profile Measurement At Optical Frequencies Using A Spectral Density Approach, John E. Nuwer, Jack Smith, Fred J. Taylor, Thomas H. Pries

LARS Symposia

The spectral properties of the intensity of a light beam, propagated in a turbulent atmosphere in the presence of wind shears, are studied. The theory of light propagation in a turbulent media is presented. Special attention is given to developing the theoretical characteristics of a spherical wave. These theoretical relations are numerically analyzed. The analysis indicate that the problem of remotely measuring winds can be automated. Several machine techniques are offered to accomplish this task.