Limitations to accuracy of land cover/use analyses with LANDSAT digital data.
dc.contributor.author | Ma, Zhenkui | en_US |
dc.contributor.advisor | Olson, Charles E., Jr. | en_US |
dc.date.accessioned | 2014-02-24T16:26:48Z | |
dc.date.available | 2014-02-24T16:26:48Z | |
dc.date.issued | 1990 | en_US |
dc.identifier.other | (UMI)AAI9023596 | en_US |
dc.identifier.uri | http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqm&rft_dat=xri:pqdiss:9023596 | en_US |
dc.identifier.uri | https://hdl.handle.net/2027.42/105270 | |
dc.description.abstract | There are difficulties in land cover/use classification of LANDSAT MSS and TM data. The minimum level of accuracy of 85 to 90 percent required for commercial application is seldom achieved with current techniques. The unsatisfactory output of digital processing of LANDSAT MSS and TM data results from several factors: non-normal distributions in populations, overlap of digital information between cover types, misrepresentation of the population with unsuitable sample data, and limitations of the classification algorithms. Analysis of these factors indicated that non-normal distributions resulted in biased estimation of population parameters when the sample size was too small to overcome distribution effects. The assumptions underlying commonly used sampling schemes were not satisfied and information from the sample data limited application of the classification algorithms. Overlap of digital information among cover types made some spectral channels less useful than expected. A measure of overlap among cover types, Brightness Value Overlapping Index (BVOI), was developed and described in mathematically. The BVOI can be used to select better subsets of data from multi-spectral information. A classification algorithm, called Nearest Spectrum, was developed in this research. The new classifier employs the correlation decision rule to maximize the application of spectral information and to overcome limitations of the data distribution when sample size is small. A measure of classification results, called Confident Accuracy, was developed to improve the degree of confidence in classification outputs. A systematic cluster sampling method was designed to gain the most information from the population. Systematic cluster sampling is simple and less expensive for obtaining information and assessing the classification accuracy. Although the accuracy of classification was unsatisfactory in some cases, LANDSAT TM data is useful for land cover/use analysis. A combination of LANDSAT TM data with other available geographic information could improve the classification of land cover. | en_US |
dc.format.extent | 201 p. | en_US |
dc.subject | Agriculture, General | en_US |
dc.subject | Agriculture, Forestry and Wildlife | en_US |
dc.subject | Environmental Sciences | en_US |
dc.subject | Remote Sensing | en_US |
dc.title | Limitations to accuracy of land cover/use analyses with LANDSAT digital data. | en_US |
dc.type | Thesis | en_US |
dc.description.thesisdegreename | PhD | en_US |
dc.description.thesisdegreediscipline | Natural Resources | en_US |
dc.description.thesisdegreegrantor | University of Michigan, Horace H. Rackham School of Graduate Studies | en_US |
dc.description.bitstreamurl | http://deepblue.lib.umich.edu/bitstream/2027.42/105270/1/9023596.pdf | |
dc.description.filedescription | Description of 9023596.pdf : Restricted to UM users only. | en_US |
dc.owningcollname | Dissertations and Theses (Ph.D. and Master's) |
Files in this item
Remediation of Harmful Language
The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.
Accessibility
If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.