Computer-Assisted Data Acceptance (CADA) Contractor Test Report.
Abstract
This report provides the first results of laboratory testing of the Computer-Assisted Data Acceptance tools for the acceptance of Contractor-provided CALS raster type 1 data. The testing to date has only been performed with legacy data obtained from the DSREDS and EDCARS sites that was converted to meet the CALS standards. It has been difficult to obtain data. Thanks to the efforts of the CALS Test Network Office (CTNO), Lawrence Livermore Laboratories, and the Communication Electronics Command (CECOM) Concurrent Engineering Directorate three set of data were obtained. Each of the three sets of contractor data were loaded, evaluated by the CADA tools and the results analyzed as reported herein. A great deal of difficulty was encountered in applying the CADA tools to the IBM and Hughes data since it was not of high quality and of a format that had not been previously analyzed. The data from IBM and Hughes consisted mainly of technical illustrations with no borders. A high rejection rate was obtained on these images because the current CADA algorithms work best with engineering drawings that conform to ANSI standards and not on technical illustrations. The ID recognition portion was not tested on these images because they have no borders. The data obtained from CECOM was from a contract with Magnavox and, of the three sets, was the most representative of the data expected to be received by the DSREDS, EDCARS, and JEDMICS repositories. About 90% of these images, consisting of engineering drawings, parts lists, and associated lists were accepted by the CADA image analysis software. However, a large percentage of these images (about 75%) were rejected in the ID recognition portion due to different reasons documented in this report.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 21, 1992
- Accession Number
- ADA312452