Invariant image classification using triple-correlation-based neural networks

see the original item page
in the repository's web site and access all digital files if the item*

1994 (EN)
Invariant image classification using triple-correlation-based neural networks (EN)

Delopoulos, Anastasios (EN)
Kollias, Stefanos (EN)
Tirakis, Andreas (EN)

N/A (EN)

Triple-correlation-based neural networks are introduced and used in this paper for invariant classification of two-dimensional gray scale images. Third-order correlations of an image are appropriately clustered, in spatial or spectral domain, to generate an equivalent image representation that is invariant with respect to translation, rotation, and dilation. An efficient implementation scheme is also proposed, which is robust to distortions, insensitive to additive noise, and classifies the original image using adequate neural network architectures applied directly to 2-D image representations. Third-order neural networks are shown to be a specific category of triple-correlation-based networks, applied either to binary or gray-scale images. A simulation study is given, which illustrates the theoretical developments, using synthetic and real image data. (EN)


Image Representation (EN)
Classification (of information) (EN)
Additive Noise (EN)
Two dimensional gray scale images (EN)
Efficient Implementation (EN)
Correlation methods (EN)
Invariance (EN)
Pattern recognition (EN)
Invariant image classification (EN)
Triple correlation based neural networks (EN)
Simulation Study (EN)
Image Classification (EN)
Neural Network (EN)
Image analysis (EN)
Algorithms (EN)
Neural networks (EN)
Computer architecture (EN)
Multilayer perceptrons (EN)
Mathematical models (EN)
Perceptron theory (EN)
Signal distortion (EN)

Εθνικό Μετσόβιο Πολυτεχνείο (EL)
National Technical University of Athens (EN)

IEEE Transactions on Neural Networks (EN)


Publ by IEEE, Piscataway, NJ, United States (EN)

*Institutions are responsible for keeping their URLs functional (digital file, item page in repository site)