What is the simplest most effective way to motion detect using webcams ?
Mostrar comentarios más antiguos
For a while now I've been searching reading and trying out some of the open source out there. With not much background in Matlab.
I'm trying to write a motion detection code based on background subtraction or temporal difference methods. And since my lack of experience when I read papers and mathematical equations I can't implement it into code and when I read source codes I can't really understand what goes on underneath.
Could you kindly point me out in the right direction. What is the best way to motion detect for a Security/ Alarm system so it needs to be fast and uninfected by noise other that intruders for closed spaces? And how can I implement it while I still understand every single line of the code?
Thank you
Respuesta aceptada
Más respuestas (2)
Image Analyst
el 13 de Abr. de 2013
0 votos
The Mathworks web site has many pages on tracking. Most are short demos that involve the Computer Vision System Toolbox or the Image Processing Toolbox or Image Acquisition Toolbox. Do a search: http://www.mathworks.com/searchresults/?q=tracking&q1=tracking&q2=&q3=¬q=&c[]=productsservices&c[]=solutions_tn
6 comentarios
Lobna
el 13 de Abr. de 2013
Editada: Walter Roberson
el 13 de Abr. de 2013
Image Analyst
el 13 de Abr. de 2013
So what's wrong with your subtraction idea? That's simple.
Lobna
el 13 de Abr. de 2013
Editada: Walter Roberson
el 13 de Abr. de 2013
Image Analyst
el 13 de Abr. de 2013
I thought I did. Did you see the link I posted? Alternatively you could search the File Exchange for tracking applications. http://www.mathworks.com/matlabcentral/fileexchange/?search_submit=fileexchange&query=tracking&term=tracking
Ankur Singh
el 31 de Mzo. de 2014
hey image analyst
in optical flow method what tells us that some motion has been detected??
Image Analyst
el 31 de Mzo. de 2014
The vector field will be non-zero where there has been motion.
tairman singh
el 18 de Abr. de 2013
try this function function [logic]=mcheck(f,s)
sub=imabsdiff(f,s);
diff_im=im2bw(sub,.07);
diff_im = medfilt2(diff_im, [3 3]);
diff_im = bwareaopen(diff_im,500);
imbw = imfill(diff_im,'holes');
[a b c]=size(imbw);
count=im2col(imbw,[a b],'sliding');
count=sum(count);
if count<=3000
logic=1;
else
logic=0;
end
end
it takes two consecutive images clicked by the camera, as the input function parameters i.e 'f' and 's' and return logic '1' or '0' if it detect motion or not. Changing different values of count ,allows you to control the sensitivity.
1 comentario
Lobna
el 22 de Abr. de 2013
Categorías
Más información sobre C4ISR en Centro de ayuda y File Exchange.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!