Weekly Report (5th July - 18th July)

Introduction

This is the report of Pallan Madhavan for the GSOC 2012 on the project "Scilab Image Processing Toolbox (SIP) " for the period 5th July to 18th July 2012.

Task List

Wrote a C interface for Dewarp function

Dewarp function helps us to dewarp a camera image based on text lines.It includes a series of process.Some of the main processes include background normalization and binarization of the given image.Then the centers of the textlines is found and a least squares quadratic fit to the longest of the line.A short, randomly-colored segments is plotted to show the initial estimate of the vertical center of each component.which are made quadratic fit over the full textline. The lines that are shorter than 0.8 of the maximum line length are not fitted.Also,The contours of constant vertical disparity, calculated from these smoothed lines, and the dewarped result on the input image after applying the vertical disparity,is then applied horizontal disparity field ,which then results a final image.

Things learnt

The concept behind dewarping an image and how leptonica can be used to do so.

Commits

b43bef1f2c9182c5abc9c84f45da066ba5fab21b
a6b19a5dca197090840e11a5714c66c9b1daba20
f211f7efc622954623c0ec176e64a4088737e73c
8417202cc955f7df911832673a58df1487b70ba5
38159e0b7f430d3789faec58a89389d056d83951

Screencast

Below are the links to screencast uploaded on Vimeo . Dewarp interface running in Scilab; https://vimeo.com/45161988