EXISTING used to compute hyperplane with maximum margin. The

EXISTING
SYSTEM:

In
the exiting system of the gender prediction from handwriting there are some
limitations. A computer is not allowed to transcript the content of difficult
handwritten document. While writing there are many difficulties are produced by
interpersonal and intrapersonal. Noisy background is also a limitation of
existing system. The supervised learning problems are considered as binary or
multi-class ones.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

 

PROPOSED
SYSTEM:

The
proposed system presents the study to predict gender of individuals from the scanned
copy of their handwritings. The proposed system is based on extracting the set
of features from writing samples of male and female writers and training
classifiers to learn to differentiates between the two. Images are differentiated
using Otsu thresholding algorithm. The following features have been considered.
Writing attributes like slant, curvature, texture and liability are estimated
by computing local and global features. Classification is done using artificial
neural network and support vector machine. Support vector machine are the sets
of related supervised learning which can be used for both classification and
regression. In simple word, an SVM classification tries to build a decision
model capable of predicting one category falls into the other. Support vector
machine classifies the images on test dataset. SVM is used to compute hyperplane
with maximum margin. The computation for the output of a given SVM,

                

 

Artificial neural network, the sequence
recognition is used for handwriting detection. The proposed technique evaluated
on the databases resulting the gender prediction from handwriting.

Some features which is used are:

·        
Tortuosity: this is used
to differential between the smooth handwriting and twisted handwriting.

·        
Direction: this is used
to measure tangent direction of text.

·        
Curvatures.

·        
Chain code.

·        
Edge direction.