Table-structure Recognition Method Consisting of Plural Neural Network Modules.
Hiroyuki AoyagiTeruhito KanazawaAtsuhiro TakasuFumito UwanoManabu OhtaPublished in: ICPRAM (2022)
Keyphrases
- neural network
- pattern recognition
- computational complexity
- high accuracy
- dynamic programming
- recognition process
- segmentation method
- unsupervised learning
- database
- significant improvement
- k means
- pairwise
- preprocessing
- object recognition
- computational cost
- recognition rate
- detection method
- computer vision
- image matching
- bp neural network
- genetic algorithm
- neural network is trained
- modular neural network