Search NASA⌕ Search

Engineering topics

Malila, W. A.

Publications and source records attributed to Malila, W. A..

At least 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
NASA NTRS1984

Study of spectral/radiometric characteristics of the thematic mapper for land use applications

The previous characterization of scan-related low-frequency noise was confirmed and extended through analysis of reflective-band data from another nighttime acquisition. Amplitude and phase relationships of the level shifts were determined for each detector in each of free full frames. Analysis of scan-direction-related signal droop effects in nighttime data from the reflective bands was begun with encouraging initial observations. Also, an effort to characterize high-frequency noise in the reflective bands through Fourier analysis of nighttime data was initiated. Recommendations are made relative to the choice of radiometric calibration constants in the thematic mapper image processing system for the routine processing of TM data. Non-linear (piece-wise linear) calibration curves are recommended.

Malila, W. A.↗
NASA NTRS1984

Using knowledge of agricultural practices to enhance through-the-season interpretation of Landsat data

Landsat data contain features that can be interpreted to produce information about crops, in support of crop estimation procedures. This paper considers ways in which detailed knowledge of agricultural practices and events might increase and improve the utilization of Landsat data in both the predictive and observational or measurement components of such procedures. Landsat observables related to agricultural practices and events throughout the cropping season are listed. Agricultural fields are identified as the preferred observational units for incorporating refined agricultural understanding, such as crop rotation patterns, into machine procedures. Uses of Landsat data from both prior seasons and the current season are considered, as is use of predictive models of crop appearance. The investigation of knowledge engineering systems tailored to through-the-season estimation problems is recommended for long range development.

Malila, W. A.↗
NASA NTRS1984

Radiometric analyses of Landsat-4 digital image data

The results of analyses of the radiometric characteristics of digital image data produced by the Landsat 4 MSS and TM sensors are reviewed. The MSS analysis found high quality data comparable to the products of previous Landsats, except for a low-level coherent noise effect having a magnitude of about 0.5 counts in each band and a spatial period of about 3.6 pixels. The TM analysis found excellent spatial resolution, generally high data quality and expected scan-angle effects. Procedures for equalizing detector responses appear to be working as intended. Two low-amplitude artifacts also were discovered. The first is associated with the direction of scan (TM employs bidirectional scanning). The second, a very low employs bidirectional scanning). The second, a very low frequency level-shift artifact which produces image banding that is most pronounced in TM1, could be a serious problem for water-related applications or others where average signal levels are low. Correction procedures are under investigation.

Malila, W. A.↗
NASA NTRS1983

Study on spectral/radiometric characteristics of the Thematic Mapper for land use applications

Previous characterization of scan-related low-frequency noise was extended and refined through detailed analysis of shutter calibration data on CCT-ADDS tapes and reflective-band data from nighttime acquisitions. A recommended correction procedure was identified that uses calibration shutter data both as a diagnostic and to obtain correction values. Through comparison of coincident TM and MSS data, illustrations of the added information content of TM data for agricultural applications were developed. The capability of improved spatial resolution to better define boundaries and to resolve spatial details is shown. Spectral analysis of tasseled-cap transformations of TM and MSS data shows high correlation between greenness features, greater signal range for TM, and indications that a subset of TM bands could accurately simulate MSS data, if required.

Malila, W. A.↗
NASA NTRS1983

Investigation of radiometric properties of the LANDSAT-4 multispectral scanner

The radiometric data quality of the LANDSAT 4 multispectral scanner (MSS) was examined using several LANDSAT 4 frames. It was found that LANDSAT 4 MSS produces high-quality data of the caliber experienced with previous LANDSATS. For example, the detector equalization procedure worked well, leaving a residual banding effect of about 0.3 digital counts RMS, close to the theoretical minimum value of quantization error. Nevertheless, artifacts of the data were found, two of which were not experienced in previous MSS data. A low-level coherent noise effect was observed in all bands, with a magnitude of about 0.5 digital counts and a frequency of approximately 28 KHz (representing a wavelength of about 3.6 pixels); a substantial increase in processing complexity would be required to reduce this artifact in the data. Also, a substantial scan-length variation (of up to six pixels) was noted in MSS data when the TM sensor was operating; the LANDSAT 4 correction algorithms being applied routinely by the EROS Data Center to produce a p-type data should remove most of this variation. Between-satellite calibrations were examined in paired LANDSAT 3 and LANDSAT 4 MSS data sets, which were closely matched in acquisition time and place. Radiometric comparisons showed that all bands were highly linear in digital counts, and a well-determined linear transformation between the MSS's was established.

Malila, W. A.↗
NASA NTRS1983

Study on spectral/radiometric characteristics of the thematic mapper for land use applications

The change in mean signal level as a function of scan angle and scan direction was studied. The overall scan angle effect corresponded to that expected based on atmospheric modeling and scene characteristics. An initial, empirical correction model employing exponential decay was developed for reflective bands. Band 6 has a significant scan direction effect which is markedly different from that found in the reflective bands. A low frequency noise was discovered which was most pronounced in Band 1, detectors 4, 12, 10, and 8, having amplitudes of approximately 2.0, 1.5, 1.0, and 0.75 quantizing levels, respectively. This low frequency variation in mean signal amplitude was highly correlated among these four Band 1 detectors. Low frequency noise was also observed in Band 7, detector 7; band 2, detector 1; Band 3, detectors 1 and 16; and Band 5, detector 10.

Malila, W. A.↗
NASA NTRS1983

Study on radiometric consistency of LANDSAT-4 multispectral scanner

Two full frames of radiometrically corrected LANDSAT-4 MSS data were examined to determine a number of radiometric properties. It was found that LANDSAT-4 MSS produces data of good quality with dynamic ranges and target responses qualitatively similar to those of previous MSS sensors. Banding appears to be quite well corrected, with a residual rms error of about 0.3 digital counts being measured; the histogram equalization algorithm appears to be working as advertised. A low level coherent noise effect was found in all bands, appearing in uniform areas as a diagonal striping pattern. The principle component of this noise was found by Fourier analysis to be a highly consistent wavelength of 3.6 pixels along a scan line (28 KHz). The magnitude of this effect ranged from about 0.75 of one count in the worst band (Band 1) to only about 0.25 counts in the best band (Band 4). Preparations were made for establishing a relative radiometric calibration from MSS 4 data with respect to MSS 3.

Malila, W. A.↗
NASA NTRS1982

Research and development of LANDSAT-based crop inventory techniques

A wide spectrum of technology pertaining to the inventory of crops using LANDSAT without in situ training data is addressed. Methods considered include Bayesian based through-the-season methods, estimation technology based on analytical profile fitting methods, and expert-based computer aided methods. Although the research was conducted using U.S. data, the adaptation of the technology to the Southern Hemisphere, especially Argentina was considered.

Horvath, R.↗
NASA NTRS1981

Development and evaluation of an automatic labeling technique for spring small grains

A labeling technique is described which seeks to associate a sampling entity with a particular crop or crop group based on similarity of growing season and temporal-spectral patterns of development. Human analyst provide contextual information, after which labeling decisions are made automatically. Results of a test of the technique on a large, multi-year data set are reported. Grain labeling accuracies are similar to those achieved by human analysis techniques, while non-grain accuracies are lower. Recommendations for improvments and implications of the test results are discussed.

Crist, E. P.↗
NASA NTRS1981

Analysis of scanner data for crop inventories

Progress and technical issues are reported in the development of corn/soybeans area estimation procedures for use on data from South America, with particular emphasis on Argentina. Aspects related to the supporting research section of the AgRISTARS Project discussed include: (1) multisegment corn/soybean estimation; (2) through the season separability of corn and soybeans within the U.S. corn belt; (3) TTS estimation; (4) insights derived from the baseline corn and soybean procedure; (5) small fields research; and (6) simulating the spectral appearance of wheat as a function of its growth and development. To assist the foreign commodity production forecasting, the performance of the baseline corn/soybean procedure was analyzed and the procedure modified. Fundamental limitations were found in the existing guidelines for discriminating these two crops. The temporal and spectral characteristics of corn and soybeans must be determined because other crops grow with them in Argentina. The state of software technology is assessed and the use of profile techniques for estimation is considered.

Horvath, R.↗
NASA NTRS1981

Analysis of scanner data for crop inventories

Accomplishments for a machine-oriented small grains labeler T&E, and for Argentina ground data collection are reported. Features of the small grains labeler include temporal-spectral profiles, which characterize continuous patterns of crop spectral development, and crop calendar shift estimation, which adjusts for planting date differences of fields within a crop type. Corn and soybean classification technology development for area estimation for foreign commodity production forecasting is reported. Presentations supporting quarterly project management reviews and a quarterly technical interchange meeting are also included.

Horvath, R.↗
NASA NTRS1980

Landsat features for agricultural applications

This paper presents relationships among Landsat MSS bands and selected transformations of them, with emphasis on the Tasseled-Cap Transformation and its Brightness, Greenness variables. It also discusses relationships between reflectance measurements made in the Landsat spectral bands and actual Landsat data. Agronomically oriented analyses of reflectance measurements of wheat throughout a growing season are presented, with a comparison of various green measures, correlation with crop development stage, and examination of the effects of moisture stress. The final example addresses the use of transformed variables in a newly developed approach to forestry change detection.

Malila, W. A.↗
NASA NTRS1980

A temporal-spectral analysis technique for vegetation applications of Landsat

This paper presents a method for describing the overall continuous pattern of crop spectral development from a set of discrete Landsat observations, using mathematical representations termed profiles. The agronomic basis for the approach is described along with model forms and techniques for estimation of the model parameters. In addition, several current or potential applications of the technology are described.

Crist, E. P.↗
NASA NTRS1980

Procedure M - A framework for stratified area estimation

This paper describes Procedure M, a systematic approach to processing multispectral scanner data for classification and acreage estimation. A general discussion of the rationale and development of the procedure is given in the context of large-area agricultural applications. Specific examples are given in the form of test results on acreage estimation of spring small grains.

Kauth, R. J.↗