Off-Line Arabic Character Recognition --- A Review

摘要: Off-line recognition requires transferring the text under consideration into an image file. This represents only available solution to bring printed materials electronic media. However, process causes system lose temporal information of that text. Other complexities off-line has deal with are lower resolution document and poor binarisation, which can contribute readability when essential features characters deleted or obscured. Recognising Arabic script presents two additional challenges: orthography is cursive letter shape context sensitive. Certain character combinations form new ligature shapes, often font-dependent. Some ligatures involve vertical stacking characters. Since not all letters connect, word boundary location becomes interesting problem, as spacing may separate words, but also certain within a word. Various techniques have been implemented achieve high rates. These tackled different aspects system. review organised five major sections, covering general overview, writing characteristics, system, OCR software conclusions.