Fourier and wavelet descriptors for shape recognition using neural networks-a comparative study

Pattern Recognition

Volumn 35, Issue 9, 2002, Pages 1949-1957

  Osowski, Stanislaw ^a   Nghia, Do Dinh ^a  

^a WARSAW UNIVERSITY OF TECHNOLOGY   (Poland)

Author keywords

Fourier and wavelet descriptors; Neural networks; Shape recognition

Indexed keywords

FEATURE EXTRACTION; FOURIER TRANSFORMS; MULTILAYER NEURAL NETWORKS; SELF ORGANIZING MAPS; WAVELET TRANSFORMS;

SHAPE RECOGNITION;

OBJECT RECOGNITION;

PATTERN RECOGNITION;

ARTIFICIAL NEURAL NETWORK; FOURIER ANALYSIS; INFORMATION PROCESSING; MATHEMATICAL ANALYSIS; RECOGNITION; SIMULATION; WAVEFORM;

EID: 0036721673     PISSN: 00313203     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0031-3203(01)00153-4     Document Type: Article

References (20)

1. A complete set of Fourier descriptors for 2-D shapes
2. Elliotic Fourier features of a closed contours
3. Wavelet descriptors for multi-resolution recognition of handprinted characters
4. Ten lectures on wavelets
5. "Neural-gas" network for vector quantization and its application to time series prediction
6. The self organising map
7. Fast second order learning algorithm for feedforward multilayer neural networks
8. Noise injection into inputs in back-propagation learning
9. Counterpropagation networks
10. Complex vectors and image identification
11. Uvi-wave