Word Spotting in Offline Handwritten Documents

In the phrase \Document Image", the word \Document" is derived from the

Latin word \documentum" which in Medieval Latin is referred to as a \written

instruction" is normally used to communicate and store information (Javed

2016). Document image processing is a subfield of digital image processing. It

mainly deals with the transformation of digitized document images into electronic

form for storage, transmission, reprocess, and modification. The objective of

document image processing is to recognize the text and graphics components in

document images, and to extract the intended information. Document image

processing can be categorized into two types such as Textual processing and

Graphics processing.

Textual processing deals with the text components of a

document image. Some tasks here are: recognizing the text by OCR, determining

the skew, finding columns, paragraphs, text lines, and words. Text document

images are composed of plain or illustrated text (books, magazines, newspaper,

archives) and structured text (forms, invoices, envelop).

Graphics processing deals with the non-textual line and

symbol components that make up line diagrams, delimiting straight lines between

text sections, company logos, etc. Graphics document images are composed of

maps, engineering drawings, and music sheets; their corresponding document

image processing goals are Geographic Information Systems (GIS) representation

for maps, Computer Aided Design (CAD) format for engineering drawing, and

Musical Instrument Digital Interface (MIDI) representation for music scores.