Title :
Neural networks to solve the problems of control and identification [automated TV-set raster tuning]
Author :
Demidenko, Serge ; Sadykhov, Raouf Kh ; Podenok, Leonid P. ; Vatkin, Maxim E. ; Klimovich, Alexey N.
Author_Institution :
Massey Univ., Palmerston North, New Zealand
Abstract :
The automated TV-set raster tuning system capable to control the vertical and horizontal size and position, pincushion, trapezoid, parallelogram and tilt has been developed. To solve the problem of non-linearity and non-symmetry approximation the neural network with activation function based on the Volterra series seas used. The new algorithm invariant to oscillation of brightness based on neocognitron neural network was applied for processing the grayscale images of integral circuits
Keywords :
Volterra series; brightness; neural nets; television receivers; tuning; Volterra series; activation function; automated TV set raster tuning system; brightness; grayscale images; horizontal size; neocognitron; nonlinearity; nonsymmetry approximation; parallelogram; pincushion; tilt; trapezoid; vertical size; Automatic control; Chebyshev approximation; Circuit optimization; Cybernetics; Distortion measurement; Informatics; Kernel; Neural networks; Nonlinear distortion; Tuning;
Conference_Titel :
Electronic Design, Test and Applications, 2002. Proceedings. The First IEEE International Workshop on
Conference_Location :
Christchurch
Print_ISBN :
0-7695-1453-7
DOI :
10.1109/DELTA.2002.994638
