Identification of fresh fuel assemblies using optical character recognition
The research involves the realization of an unattended measurement station (UMS) for fresh fuel assemblies, The station is composed of two computers, a hose personal computer (PC) in charge of the characterization of the radioactive properties of the fuel element (FE) and a second PC for the FE automated identification. A video image of the identification characters engraved onto the FE is used as input for the automated identification procedure. The image is processed and the connected components abiding to given criteria are extracted. An optical character recognition (OCR) procedure analyses this information and outputs the FE identifications in ASCII format, which is then transmitted to the host computer to be inserted in the FE record. For each character in the identification string, the OCR outputs a confidence index, which is used by the system to require specific confirmation from the inspector during periodic visits. The inspector will then view the original image and make the necessary edition. The FE identification system, its architecture at the software and hardware levels are described. The operating principles of the optical character recognition system are also described as well as the communication protocol between the two computers. Finally an appreciation is given of the first field tests of the UMS at the Fabbricazione Nucleare facility at Bosco Marengo.
Bibliographic Reference: Paper presented: 19th Annual Symposium on Safeguards and Nuclear Material Management, Montpellier (FR), May 13-15, 1997
Availability: Available from (1) as Paper EN 40600 ORA
Record Number: 199710856 / Last updated on: 1997-07-08
Original language: en
Available languages: en