Automatic Numeral Recognition System Using Local Statistical and Geometrical Features

Optical Character Recognition (OCR) research includes computer vision, artificial intelligence, and pattern recognition. recognition has garnered a lot of attention in the last decade due to its broad variety uses applications, including multiple-choice test data, business documents (e.g., ID cards, bank notes, passports, etc.), automatic number plate This paper introduces an system for printed numerals. The reading is based on extracting local statistical geometrical features from text image. Those are represented by eight vectors extracted each digit. Two these (A, A th), six (P1, P2, P3, P4, P5, P6). Thus, database created consists 1120 features. For purpose recognition, image compared with all images saved depending value Minimum Distance (MD). All digits (0–9) were identified 100% accuracy. average computational time required recognize numeral at any font size 0.06879 seconds.