okralact - a multi-engine Open Source OCR training system.
Konstantin BaiererRui DongClemens NeudeckerPublished in: HIP@ICDAR (2019)
Keyphrases
- open source
- open source software
- post processing
- training set
- source code
- online learning
- document images
- error correction
- preprocessing
- character recognition
- test set
- data sets
- feedforward neural networks
- optical character recognition
- training phase
- training algorithm
- training samples
- digital libraries
- real world
- training examples
- feature space
- serious games
- training process
- feature extraction
- decision trees
- neural network
- database
- document image analysis