Title :
In-camera detection of fabric defects
Author :
Baykal, Ibrahim Cem ; Jullien, G.A.
Author_Institution :
Dept. of Electr. & Comput. Eng., Calgary Univ., Alta., Canada
Abstract :
Industrial inspection cameras for Web processes have a very high output video rate. This output data rate can be reduced by preprocessing the video stream inside the camera in order to only send pixels that have a high probability of containing defect information. Co-occurrence matrices are quite successful in detecting structural defects in fabrics; however, they require high processing power. This paper discusses simplification to these algorithms, which are simple enough to fit into a medium-small size FPGA.
Keywords :
cameras; fabrics; field programmable gate arrays; image texture; inspection; matrix algebra; Web processes; co-occurrence matrices; data rate; defect information; fabric defects; in-camera detection; industrial inspection cameras; medium-small size FPGA; processing power; structural defects; video rate; video stream preprocessing; Cameras; Computer industry; Electrical equipment industry; Electronics industry; Fabrics; Field programmable gate arrays; Industrial control; Inspection; Machinery production industries; Streaming media;
Conference_Titel :
Circuits and Systems, 2004. ISCAS '04. Proceedings of the 2004 International Symposium on
Print_ISBN :
0-7803-8251-X
DOI :
10.1109/ISCAS.2004.1328901
