Ask Your Question

insaf's profile - activity

2012-11-09 10:32:20 -0500 answered a question Classification of object from a video/Image ( human, animals, others(cars etc.,))

Hi,

I have to do the same for my project. We can talk more about that if you want : insaf.ist@gmail.com

See you.

2012-11-09 06:48:35 -0500 answered a question How to use haartraining in Video object classification?

Hello, To classify the content of the video you have to make some prior processing. First you have to make a background subtraction, you'll have a mask of your moving objects, then you have to translate them to a list. After that, you have to track your objects (this is not always necessary). By having all this in hand, you can classify your objects using haar features or whatever you want.

I can help you on the background and tracking, however, I am not really confident in that (I myself need some help in understanding code I already have). I have also to classify objects in videos and I am working on that.

Keep in touch if any solutions you have and I am as well available if any questions.

2012-10-30 02:25:47 -0500 asked a question Exercice 6 of chapter 5 of "learning opencv computer vision with the opencv library"

Hello everybody,

I am new in the wiki and also new in open cv.

I wish you could help me in my issue. I want to do exercice 6 chapter 5 of the book "learning opencv computer vision with the opencv library" but I don't know how. What is asked in the exercice is to flood fill the white part of the image with the value 100, so what I did, I cheked for every pixel if it is white, and then flood fill, but I don't think this is the way, here's a part of my code:

for( int y=0; y<diff13->height; y++ ) { uchar* ptr = (uchar) ( diff13->imageData + y * diff13->widthStep ); for( int x=0; x<diff13->width; x++ ) { if(ptr[3x+1] == 255) {
ptr[3x+1] = 100; cvFloodFill(diff13,cvPoint(ptr[3x+1],y),color2, cvScalarAll(5.0), cvScalarAll(5.0), cc, 4, NULL ); } } } Can anyone help me please ??? I can provide more details.

Thanks a lot, Insaf.