Document Reader Using An Android Platform |
Author(s): |
| Bhagyashri Jadhav , DCOER,pune; Pratiksha Hundare, DCOER,pune; Tejashree Jadhav, DCOER,pune; Sonali Jagtap, DCOER,pune |
Keywords: |
| Put your keywords here, keywords are separated by comma |
Abstract |
|
Our purpose was to create an application that allows a user to jump to the correct section of an audio book by taking a picture of the page they are reading in a paper book. We used image processing algorithms such as the Hough transform to rotate pictures of the page, the Otsu’s method is used to threshold and binarize the image and brick wall coding (BWC) to detect features on the page. Then We will used a k-nearest-neighbour algorithm to assign a score to a test page by comparing it with pages in our database. We returned the closest matching page to the user generated picture. We obtained 40% accuracy over a dataset of 20 pages. In future work we could improve our accuracy by using angles between features which are more invariant to the pictures the user takes. |
Other Details |
|
Paper ID: IJSRDV2I11063 Published in: Volume : 2, Issue : 11 Publication Date: 01/02/2015 Page(s): 135-136 |
Article Preview |
|
|
|
|
