User:TheLarry/Notebook/Larrys Notebook/2010/03/02

Tracking

Main project page Previous entry      Next entry

Simulating Images This is the first microtubule i have created using the method of placing airy disks in a line. I have not created a larger image yet. So there is no sub pixel movement. Resampling to a lower res image is next. And after that adding Gaussian Noise. but at least i am on my way. Resampling was easy. The last time i did this, I set this lower resoluted image into an image that matched the size of a .png from the camera (658x479 (or something liek that)). but i am thinking that isn't necessary. so i want to hold off doing it unless Koch says it is important. If he does i'll just use the sub.vi IMAQ ImageToImage. But for now i'll hold off The good things about this method compared to the last time i had to do this: No fourier space So no complex plane or anything having to deal with that Don't gotta add Poisson Noise since it is already there from the random number picks I haven't double checked this but i think it makes sense Keeping track of a handful of dye locations was easier than keeping track of the entire length of the microtubule And as for now no resampling Now i have to add the Gaussian Noise background OK i think i have a handle on making these images. seen below is a frame from the first movie i have saved from the software. I'll track it tomorrow. I don't have a handle on these float images like how bright they can get and so on. but i will. Anyway the software is working pretty well which i am happy about. It is still slow from the random number generating but if there is a way past that i'll find it. The steps are to make a shitload of movies and track it with the tracking software. That will go into the paper to show how cool this software is. Koch asked me to write up this image generation well so i'll take the time and do that now. /Image Generating Software