A novel syntactic approach is introduced to treat particular problems in pattern recognition. The procedure is implemented by the use of optical correlation methods for identifying the various primitives that appear in the input pattern, and their importance is determined by fuzzy relational scoring. Robust pattern recognition with tolerance to normal variations is demonstrated, indicating an efficient new approach for optical pattern recognition.
© 1996 Optical Society of America
Rajani Srinivasan, Jason Kinser, Marius Schamschula, Joseph Shamir, and H. John Caulfield, "Optical syntactic pattern recognition by fuzzy scoring," Opt. Lett. 21, 815-817 (1996)