Original scientific paper

Character recognition based on region pixel concentration for license plate identification

Krešimir Romić ; Faculty of Electrical Engineering, Josip Juraj Strossmayer University of Osijek, Kneza Trpimira 2B, HR-31000 Osijek, Croatia
Irena Galić ; Faculty of Electrical Engineering, Josip Juraj Strossmayer University of Osijek, Kneza Trpimira 2B, HR-31000 Osijek, Croatia
Alfonzo Baumgartner ; Faculty of Electrical Engineering, Josip Juraj Strossmayer University of Osijek, Kneza Trpimira 2B, HR-31000 Osijek, Croatia

Abstract

Since it has become possible to perform digital image processing in a short period of time, its usage in technical systems is getting more common. Automatic license plate recognition is one such example. By using digital image processing it is possible to automatically detect and recognize characters on vehicle license plates. The steps taken in this process are image preprocessing, plate detection, character segmentation and recognition. This process is performed by the algorithm which takes the digital image as an input and gives textual form of license plate characters as an output. There are several methods used to perform this process. These methods are explained in this paper and one of them is implemented in C# programming language. The results show that the algorithm works fine in cases without much deformation of an input image. However, there are still cases where unpredictable nature of an input image can cause unsuccessful detection or recognition.

Keywords

character recognition; detection; digital image processing; region pixel concentration; segmentation

Hrčak ID:

83906

URI

https://hrcak.srce.hr/83906

Publication date:

29.6.2012.

Article data in other languages: croatian

Visits: 5.331 *