The classification step classifies the image of each digit using the features extracted in the previous step. These may include a histogram of oriented gradients (HOG), intensity values, or Gabor features. The feature extraction step attempts to extract the most useful features of a single digit’s image in order to improve character classification. One may divide this step into the two sub-steps of feature extraction and classification as was done in a previous investigation. Many methods for character recognition have been investigated. The second step is character recognition in which the segmented characters are identified.
However, it has been reported that individual character classification usually requires longer computation times than simultaneous classification of all digits. To the best of our knowledge, all reported methods execute single digit classification after segmentation of each digit. After extracting a region for the serial number, segmentation of the serial number is implemented to classify each digit individually. To resolve these problems, several methods have been investigated using image processing techniques. ROI detection is challenging for various reasons including uneven illumination, smears, and patterns present in the background. The first step includes detection of the region of interest (ROI) that contains the serial number in the acquired banknote image as well as enhancement of the acquired image. Typical serial number recognition systems execute a pre-processing step and a character recognition step. Consequently, banknote serial number recognition requires the design of a fast serial number recognition system with very high accuracy, a challenging task. In addition, real-time recognition must meet the required transaction times for teller machines. Unlike these other fields, serial number recognition for banknotes requires extremely high accuracy since even a small error can result in a huge financial loss. The serial number recognition problem essentially belongs to the printed character recognition problem which has been studied in other fields such as license plate recognition and address recognition. For this reason, many investigations of serial number recognition have been conducted. Correct recognition of banknote serial numbers is important because it can be used to trace the circulation routes of individual banknotes to provide essential information about economic activities.
0 kommentar(er)
