BROWSE
Detail
Conference Paper
Assessment of the Classification Capability of Prediction Models for Postal Address Recognition in Unmanned Postal Acceptance/Delivery System
Cited - time in 
Share 
Share
- Authors
- Moon-Sung Park, Ju-Wan Kim, Jae-Gwan Song
- Issue Date
- 2015-07
- Citation
- International Conference on Frontiers of Information Technology, Applications and Tools (FITAT) 2015, pp.105-109
- Language
- English
- Type
- Conference Paper
- Abstract
- The main objective of our study is to evaluate the classification capabilities of prediction models and propose which are most likely to be suitable for postal recognition parameter setting. We also develop and then suggest a novel methodology useful in developing the various features of postal items image helpful in recognizing address. Various prediction models are applied in order to detect and extract those which provide the better accuracy in postal items image data. In our experiments, postal items image features are used for building the prediction model. The prediction models are function-based prediction and fuzzy-based method. As a result, function-based (SVM and ANNs) algorithms outperformed the other prediction models.
- KSP Keywords
- Address recognition, Delivery systems, Fuzzy-based method, Image Features, Image data, Parameter setting, function-based, prediction model