We have some scanned images ... , which will have the grid of cells , each for one character.
we have to fetch each characters image and store it in specified folder.
1. the scan image will not be straight. so first we have to make it straight ( this will be the biggest part in programming )
2. then crop the image to outer blue boundary of table ( see image) and scale up or down to standard size.
3. now we have standard size image . so we just have to crop small parts of the image and save it as image file for individual character.
for scaling and cropping image together . you can use approach that ,
for each border of image ( i.e top , bottom , left and right )
( we are considerinf left side in this loop )
removing a singe row or column ( by using cropping ) from that sides until you encounter a blue colour in any of the pixel in that row or coulmn
once you encounter a pixel , check if its in upper half of that side or lower half . if its in upper half rotate image in clockwise direction . if in lower half rotate in anticlock wise direction.
again start processing that side and remove column untill you find borders blue color .
repeat this till you find that first column is having all blue color pixels ...
repeat this for all sides
this will give you image which is having nothing but just a table in its 100% area. then you have to scal and get the character wise image and store it in folder
see sample images below.
please contact for clarification.
we will not consider bids above $100
Hi all bidders ..
some clarification about project
Our requirements are changed a bit... ( rather I have added a bit more work )
We need to make a simple GUI with two main screens ...
first will accept the scanned image and do the image processing as above .. and store the images in folder
the second screen is simple merging of images to construct image that has handwritten fonts images ... for the specified text ) ...
second screen will accept text in a text area. and will build the corresponding image of the text ... ( basically it merges the images of folder specified by user into one single large image )
( text to handwritten fonts )
no resizing of font needed . the output image will be of fixed size.
( you have to take care that text dont overflow out of the fixed image size , that is you have to go to next line of 1st line is over.) and support for "space" and newline
I want this to be made in 4-5 days
when saving images to folder from scanned image we have to convert is to black and white image. and only the part of character which inside the boundary should be captured ( not the corresponding computer printed font .. )
8 freelancers are bidding on average $141 for this job
Hi, I'm very much interested and serious to do this project. I have done image manipulation before and I think I suite to this project. Hope you can consider my bid. Regards