Interaction of electromagnetic energy with agricultural crops
Studying interaction of electromagnetic radiation with plants, soils, and water, and wavelengths and procedures for discriminating agricultural crops
SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Studying interaction of electromagnetic radiation with plants, soils, and water, and wavelengths and procedures for discriminating agricultural crops
Crop and soil identification from aerial photography
Crop species and soil condition computerized identification from film optical density differences, using multibase and multiemulsion photography
Crop surveys from multiband and multibase satellite photography during Apollo 9 mission, using statistical multispectral pattern recognition digital techniques
Semi-automatic crop inventory from sequential ERTS-1 imagery for Imperial Valley, California
Crop identification using ERTS-1 imagery for southeastern South Dakota
Application of ERTS-1 data to analysis of agricultural crops and forests in Michigan
Crop species recognition and mensuration in Sacramento Valley from ERTS-1 data
Identification of agricultural crops in Dekalb, Ogle, and Lee counties in Illinois by computer processing of ERTS-1 MSS data
The results reported are based on analysis of ERTS Frame 1033-15580 collected over southwestern Lower Michigan on August 25, 1972. Major agricultural crops such as corn and soybeans were approaching maturity at this data and forest canopies were dense. Extensive ground truth information was gathered by detailed field study of test strips. This detailed information was supplemented over larger areas by interpretation of RB-57 and C-47 photography and MSS imagery. Recognition processing of ERTS-1 MSS data was carried out on a digital computer. Fields and forest stands were selected as training sets and test areas. Aerial imagery was essential for locating the positions of these selected areas on ERTS digital tapes. The recognition process was successful for each type of vegetation which had a dense green canopy such as forests, corn, and soybeans. Bare soil was also recognizable as a category.
The need for multispectral data processing methods to permit the estimation of proportions of objects and materials appearing within the instantaneous field of view of a scanning system is discussed. An algorithm developed for proportion estimation is described as well as other supporting processing techniques. Application of this algorithm to space-simulated multispectral scanner data is discussed and some results presented and compared. Results indicate that, for this data set, the true proportions of the various crops contained within this data set are with one exception more closely in agreement with the proportions determined by the proportion estimation algorithm than with the proportions determined by conventional classfication algorithm.
The author has identified the following significant results. The crop types of a Great Plains study area were mapped from color infrared aerial photography. Each field was positively identified from field checks in the area. Enlarged (50x) density contour maps were constructed from three ERTS-1 images taken in the summer of 1973. The map interpreted from the aerial photography was compared to the density contour maps and the accuracy of the ERTS-1 density contour map interpretations were determined. Changes in the vegetation during the growing season and harvest periods were detectable on the ERTS-1 imagery. Density contouring aids in the detection of such charges.
The author has identified the following significant results. Multispectral scanner data obtained by ERTS-1 over six test sites in the Central United States were analyzed and interpreted. ERTS-1 data for some of the test sites were geometrically corrected and temporally overlayed. Computer-implemented pattern recognition techniques were used in the analysis of all multispectral data. These techniques were used to evaluate ERTS-1 data as a tool for soil survey. Geology maps and land use inventories were prepared by digital analysis of multispectral data. Identification and mapping of crop species and rangelands were achieved throught the analysis of 1972 and 1973 ERTS-1 data. Multiple dates of ERTS-1 data were examined to determine the variation with time of the areal extent of surface water resources on the Southern Great Plain.
Transonic flutter characteristics of three geometrically similar delta-wing models were experimentally determined in the Langley transonic dynamics tunnel at Mach numbers from about 0.6 to 1.2. The models were designed to be simplified versions of an early supersonic transport wing design. The model was an aspect-ratio-1.28 cropped-delta wing with a leadingedge sweep of 50.5 deg. The flutter characteristics obtained for this wing configuration indicated a minimum flutter-speed index near a Mach number of 0.92 and a transonic compressibility dip amounting to about a 27-percent decrease in the flutter-speed index relative to the value at a Mach number of 0.6. Analytical studies were performed for one wing model at Mach numbers of 0.6, 0.7, 0.8, and 0.9 by using both doublet-lattice and lifting-surface (kernel-function) unsteady aerodynamic theory. A comparison of the analytical and experimental flutter results showed good agreement at all Mach numbers investigated.
ERTS-1 data, obtained during the period 25 August 1972 to 5 September 1973 over a range of test sites in the Central United States, have been used for identifying and mapping differences in soil patterns, species and conditions of cultivated crops, and conditions of rangelands. Multispectral scanner data from multiple ERTS passes over certain test sites have provided the opportunity to study temporal changes in the scene. Multispectral classifications delineating soils boundaries in different test sites compared well with existing soil association maps prepared by conventional means. Spectral analysis of ERTS data was used to identify, maps, and make areal measurements of wheat in western Kansas. Multispectral analysis of ERTS-1 data provided patterns in rangelands which can be related to soils differences, range management practices, and the extent of infestation of grasslands by mesquite (prosopis fuliflora) and juniper (juniperus spp.).
The value of improving information for forecasting future crop harvests was investigated. Emphasis was placed upon establishing practical evaluation procedures firmly based in economic theory. The analysis was applied to the case of U.S. domestic wheat consumption. Estimates for a cost of storage function and a demand function for wheat were calculated. A model of market determinations of wheat inventories was developed for inventory adjustment. The carry-over horizon is computed by the solution of a nonlinear programming problem, and related variables such as spot and future price at each stage are determined. The model is adaptable to other markets. Results are shown to depend critically on the accuracy of current and proposed measurement techniques. The quantitative results are presented parametrically, in terms of various possible values of current and future accuracies.
An evaluation is conducted of the relative effectiveness of the computer analysis techniques which are commonly employed to extract land use (crop identification) information from digitized aerial photographs. It is found that the minimum distance to the mean (MDM) algorithm and the maximum likelihood ratio (MLR) can both be used for the successful recognition of land-use patterns. The MDM algorithm is slightly more accurate in cases involving the use of three or more variables. The use of the MLR algorithm, however, is preferable in cases in which less than three variables are employed.
The results of classifications and experiments performed for the Crop Identification Technology Assessment for Remote Sensing (CITARS) project are summarized. Fifteen data sets were classified using two analysis procedures. One procedure used class weights while the other assumed equal probabilities of occurrence for all classes. In addition, 20 data sets were classified using training statistics from another segment or date. The results of both the local and non-local classifications in terms of classification and proportion estimation are presented. Several additional experiments are described which were performed to provide additional understanding of the CITARS results. These experiments investigated alternative analysis procedures, training set selection and size, effects of multitemporal registration, the spectral discriminability of corn, soybeans, and other, and analysis of aircraft multispectral data.