02-10-2012 12:25 PM
My final year project is "Text Input System developed by Lips Image Recognition based on Labview for Serious Disabled".
In this, image of person's face will be acquired by CCD camera.. Then it has stages like face tracking, lips area allocation and extraction and further processing.
Then the status of mouth-open or mouth-close will be acquired in binary format as 1 & 0 respectively..This information will be given to MORSE CODE TEXT INPUT SYSTEM which will convert morse code into english text...I am having problem in developing program to track face and allocate lips area...Kindly help...
02-13-2012 08:29 AM
Post your developed vi with sample images. So that can help you better to resolve your problem.
02-14-2012 09:02 AM
02-17-2012 08:17 AM
Do you have any images with you. If so post it. Am out of my office so i cant check it out with cam right now.
02-20-2012 08:59 AM
Sorry for the delay. You would like to track the face and lips right?. Are you tried with pattern matching algorithm to do the same creating a template like this. At first track the eyes and nose or else anyone is enough. With referrence to that find the lips position.
Then regarding the open/close position of the lips can be done with particle analysis. Post some image for the same.
02-21-2012 12:29 AM - edited 02-21-2012 12:30 AM
As you mentioned about the allocating lips area with respect to the position of nose, can it be done using ROI...? if so then how can we determine the ROI of lips...? Also when the person moves the head, then will the ROI change..?
02-21-2012 12:56 AM
Also when the person moves the head, then will the ROI change..?
You can fix the ROI by setting a cordinate system in your image. Eventhough Person moves the head your roi will also moves with respect to the coordinate system.