Fuzzy set methods for object recognition in space applications

Keller, James M.

Progress on the following tasks is reported: (1) fuzzy set-based decision making methodologies; (2) feature calculation; (3) clustering for curve and surface fitting; and (4) acquisition of images. The general structure for networks based on fuzzy set connectives which are being used for information fusion and decision making in space applications is described. The structure and training techniques for such networks consisting of generalized means and gamma-operators are described. The use of other hybrid operators in multicriteria decision making is currently being examined. Numerous classical features on image regions such as gray level statistics, edge and curve primitives, texture measures from cooccurrance matrix, and size and shape parameters were implemented. Several fractal geometric features which may have a considerable impact on characterizing cluttered background, such as clouds, dense star patterns, or some planetary surfaces, were used. A new approach to a fuzzy C-shell algorithm is addressed. NASA personnel are in the process of acquiring suitable simulation data and hopefully videotaped actual shuttle imagery. Photographs have been digitized to use in the algorithms. Also, a model of the shuttle was assembled and a mechanism to orient this model in 3-D to digitize for experiments on pose estimation is being constructed.

Document ID

19930007370

Acquisition Source

Legacy CDMS

Document Type

Contractor Report (CR)

Authors

Date Acquired

September 6, 2013

Publication Date

September 30, 1991

Subject Category

Report/Patent Number

Accession Number

93N16559

Funding Number(s)

Distribution Limits

Public

Work of the US Gov. Public Use Permitted.

Available Downloads

Name

Type

19930007370.pdf

STI

No Preview Available

NTRS

NTRS - NASA Technical Reports Server

Available Downloads

Related Records