03-03-2011, 03:43 PM
presented by:
Chandan Kumar
Tanvi Afroze
Nibedita Barman
Satyajit Das
[attachment=9466]
Character recognition in a hand written document image
objectives
The main objective of our project is to facilitate character segmentation in handwritten document image documents. A hand written document image consists of numerous lines of hand written text
These lines are independent of any attributes like font, size etc. So, through segmentation techniques we have to differentiate letters among words and this process continues for the whole document image. e.g…
Methodologies
Now, our next step is to find the rows in the matrix which are containing at least one 1’s which indicates that this row is a part of the text.
Results and Discussions
We will also try to locate the cluster of totally white rows(lines) i.e all the cells are 0’s .These totally white rows will assist us in finding the separate sentences in the text of a handwritten document image. Consequently, we will get cluster of rows say (Row no.(1,2,3 ),(12,13,14,15),(24,25,26,27)) which are totally white.
Till now ,we have been able to read the input image, apply thinning morphology on it and write the output file on a new image file.
Thinning is a morphological operation that is used to remove selected foreground pixels from binary images, somewhat like erosion or opening. It can be used for several applications, but is particularly useful for skeletonization. In this mode it is commonly used to tidy up the output of edge detectors by reducing all lines to single pixel thickness. Thinning is normally only applied to binary images, and it gives binary image as output. e.g….
Thinning operation has been carried out with the help of structuring elements which are themselves 2-D matrices which are checked through iterative rotation process. We check whether a portion of image matrix can totally contain that structuring element or not. If yes, then we keep those pixels in our proposed output matrix, otherwise we ignore those pixels. This process goes on till we reach the penultimate column and row of image matrix.
Scope of future work
There is a lot of scope of improvement and optimization in our proposed methodologies and tools which can eventually lead to a much convenient approach towards character recognition in hand written image document. Some of the visible and worth mentioning ones are
Conclusion
Till now , in our 7th semester we have successfully completed the connected component analysis and thinning operation on a the hand written document image. In the next semester, we aim to fulfill the objective of our project i.e. character recognition in a hand written document image. We are keen to accomplish the features and goals mentioned in future scope of our project.
