20-08-2012, 04:21 PM
HANDWRITING RECOGNITION
HANDWRITING RECOGNITION.docx (Size: 521.12 KB / Downloads: 36)
ABSTRACT
Recognition involves sensing user input or other information about the user, and interpreting it. Some traditional forms of recognition include speech, handwriting, and gesture recognition.
Other types of recognition include face recognition, activity recognition (an important component of context-aware computing.
Pattern recognition consists of two stages: feature extraction and classification. Feature extraction is the measurement on population entities that will be classified.
This assists classification phase by looking for features that fairly allows to distinguish between different classes.
INTRODUCTION
It is a challenging issue to develop a practical cursive, handwritten CR system which can maintain high recognition accuracy and is independent of the quality of the input documents. Very often adjacent characters tend to be touched or overlapped. Therefore, in the segmentation-based strategy, it is essential to segment a given string correctly into its character components. The complexity of character segmentation stems from the wide variety of fonts, rapidly expanding text styles and poor image characteristics. Touched, overlapped, separated, and broken characters are major factors for causing segmentation errors. In most of the existing segmentation algorithms, human writing is evaluated empirically to deduce rules. Sometimes the rules derived are satisfactory but there is no guarantee for their optimum results in all styles of writing. Moreover human writing varies from person to person and even for the same person depending on mood, speed, environment etc. On the other hand researchers have employed techniques like articial neural networks, hidden Markov models and statistical classifiers to extract rules based on numerical data.
PROBLEM DEFINITION
Handwriting recognition is the ability of a computer to receive and interpreted intelligible handwritten input from sources such as paper documents, photographs, touch-screens and other devices. The problem of handwriting recognition can be classified into two main groups, namely off-line and on-line recognition, according to the format of handwriting inputs.
The off-line recognition systems recognise the characters after they have been written on a piece of paper, scanned into the computer and stored in the image format. On the otherhand, the on-line systems can access dynamic information of handwriting strokes while the characters are being written on a tablet or a digitiser. Although the recognition processes of both systems are different in their details, they can be broken down into five fundamental sequential stages: pre-processing, segmentation, feature extraction, classification and post-processing.
Input: Samples are read to the system through a scanner.
Preprocessing: Preprocessing converts the image into a form suitable for subsequent processing and feature extraction.
Segmentation: The most basic step in CR is to segment the input image into individual glyphs. This step separates out sentences from text and subsequently words and letters from sentences.
Feature extraction: Extraction of features of a character forms a vital part of the recognition process. Feature extraction captures the vital details of a character.
Classification: During classification, a character is placed in the appropriate class to which it belongs.
Post Processing: Combining the CR techniques either in parallel or series.
TYPES OF HANDWRITING RECOGNITION:
The online handwriting recognition problem has a number of distinguishing features which must be exploited to get more accurate results than the online recognition problem :
It is adaptive-The immediate feedback is given by the writer whose corrections can be used to further train the recognizer
It is a real time process- It captures the temporal or dynamic information of the Writing.
loops and cusps are easier and faster with the pen trajectory data than on pixel images
Segmentation is easy- Segmentation operations are facilitated by
This information consists of the number of pen strokes i.e the writing from pen down to pen up- the order of pen strokes the direction of the writing for each pen stroke and the speed of the writing within each pen stroke.
Very little preprocessing is required- The operations such as smoothing deslanting deskewing and feature extraction operations such as the detection of line orientations corners using the pen lift information particularly for handprinted characters.
Ambiguity is minimal- The discrimination between optically ambiguous characters may be facilitated with the pen trajectory information
It captures human signatures and sketches
On the other hand the disadvantages of the online character recognition are as follows:
The writer requires a special equipment which is not as comfortable and natural to use as pen and paper
It can not be applied to documents printed or written on papers
Punching is much faster and easier than handwriting for small size alphabet such as English or Arabic
The major advantage of the offline recognizers is to allow the previously written and printed texts to be processed and recognized Some applications of the online recognition are large scale data processing such as postal address reading cheque sorting short hand transcription automatic inspection and identification reading aid for visually handicapped
The drawbacks of the offline recognizers compared to online recognizers are summarized as follows-
Offline conversion usually requires costly and imperfect
pre processing techniques prior to feature extraction and recognition stage
They do not carry temporal or dynamic information such as the number and order of pen on and pen off movements the direction and speed of writing and in some cases the pressure applied while writing a character
They are not real time recognizers
ILLUSTRATIONS ON EXISTING RESEARCHES
Pre-processing is essential to prepare handwriting inputs to be suitable for later recognition stages. Examples of the pre-processing stage include the processes to identify text regions within the handwritten documents, eliminate imperfections, and segment the texts into isolated characters.
Segmentation is an important step in image processing. It is used for partitioning or separation of words ,lines or characters.
The goal of feature extraction stage is to obtain a compact description (a feature vector) that can be used to uniquely represent the character.
Classification is the main decision making stage in which the extracted features are classified into one of several categories. Modern handwriting recognition research isdominated by the use of statistical methods for classification, i.e. statistical classifiers
Post-processing typically forms a verification step, such as the use of language models and contextual information to verify the recognised characters or words.