Improving OCR performance using character degradation models and boosting algorithm
Introduction
In this paper, we study the effectiveness of a boosting algorithm (Drucker et al., 1993) in improving the performance of OCR. The original theoretical work on the boosting algorithm was done by Schapire (1990). He showed that it is in principle possible for a combination of weak classifiers (whose performances are a little better than random guessing) to achieve an arbitrarily low error (on the training data set). Drucker et al. (1993) applied the boosting algorithm to character recognition. They produced a large number of training patterns by deforming the original character images by various degrees. It was shown that the performance of character recognition was dramatically improved over that of the single network which was used as the first network in the boosting hierarchy. However, it remains to be answered whether the boosting ensemble outperforms the standard ensemble of independently trained networks. In this paper, we provide a comparative study of the boosting ensemble and the standard ensemble. We also introduce three character degradation models in the boosting algorithm.
Section snippets
Boosting algorithm
In the boosting algorithm, the weak classifiers are trained hierarchically to learn harder and harder parts of a classification problem. The algorithm requires an oracle to produce a large number of independent training patterns. The basic boosting algorithm works as follows.
- 1.
Generate a set of training data and train the first classifier.
- 2.
Generate a set of training data for training the second classifier in the following manner: Flip a coin. If it heads up, the oracle generates a pattern and
Character degradation models
We introduce three document degradation models in the boosting algorithm: (i) affine transformation, (ii) an image deformation model used in (Jain et al., 1996), and (iii) a probabilistic model for document degradation (Kanungo, 1996).
The affine model is a linear transformation of coordinate systems which take into consideration the following operations: (i) translation, (ii) scaling, (iii) rotation, and (iv) shearing. In our OCR system, character features are invariant to translation and
Experiments and discussions
We used the lower-case alphabets of the NIST (National Institute of Standards and Technology) Special Database 3 (SD3: 39,636 segmented characters) and Test Data 1 (TD1: 12,000 segmented characters); these databases consist of pre-segmented characters used in the 1992 comparative study (Wilkinson et al., 1992). In our experiments, the SD3 data set was further partitioned into a training data set (SD3-train) with 27,636 characters and a validation data set (SD3-valid) with 12,000 characters.
The
Conclusions
We have introduced three character degradation models in the boosting training. We compare the boosting ensemble with the standard ensemble of networks trained independently with character degradation models. Both the ensembles outperform the single network trained using the character degradation models. An interesting discovery in our comparison is that although the boosting ensemble has a slightly higher accuracy than the standard ensemble at zero reject rate, the advantage of the boosting
References (6)
- Drucker, H., Schapire, R., Simard, P., 1993. Improving performance in neural networks using boosting algorithm. In:...
- Jain, A., Zhong, Y., Lakshmanan, S., 1996. Object matching using deformable templates. IEEE Trans. Pattern Anal....
- Kanungo, T., 1996. Document degradations models and a methodology for degradation model validation. Ph.D. Thesis....
Cited by (21)
Efficient character segmentation approach for machine-typed documents
2017, Expert Systems with ApplicationsCitation Excerpt :Diverse approaches for character segmentation have been presented in the past (Casey & Lecolinet, 1996 ). Usually, the process of character segmentation and its importance in complex systems, such as OCR systems (Bourbakis, Pereira, & Mertoguno, 1996; Grafmüller & Beyerer, 2013; Mao & Mohiuddin, 1997; Vamvakas, Gatos, Stamatopoulos, & Perantonis, 2008), is unduly underestimated (Lu, 1995; Lu & Shridhar, 1996). Related works can be classified based on a couple of characteristics.
Recognition of handwritten Chinese characters by critical region analysis
2010, Pattern RecognitionApplication of Dempster-Shafer theory in condition monitoring applications: A case study
2001, Pattern Recognition LettersOffline handwriting recognition using synthetic training data produced by means of a geometrical distortion model
2004, International Journal of Pattern Recognition and Artificial IntelligenceIntelligent technologies in manufacturing cyber-physical systems: Evolution and future development
2023, Advanced Signal Processing for Industry 4.0Research and development of neural network ensembles: a survey
2018, Artificial Intelligence Review