Philip's profile - activity

I found that I can build it by just following what you did then remove any project which fails to compile. In my case, right now, only one project failed to compile so I had to remove it and its two test projects.

Can you tell me why?

I have a byte buffer ( unsigned char * ) of RGBA format, I wish to use this image data in OpenCV, use it as a IplImage. How to convert from this byte buffer to IplImage?

Think I have worked this out myself, still testing.

IplImage* img = cvCreateImage( cvSize(width,height), IPL_DEPTH_8U, 4);
cvSetData(img, buffer, width*4);

Has anyone found any source code for Pealing?

Hi, I have some text on a gradient background, this is a PNG image captured from a HTML rendered webpage of text. The background of this page is a gradient.

How can I go about removing the gradient background from the image and retain the text?

There is aliasing between the font and the background, would this approach work better due to the aliasing being mixed colors between the foreground and background?

I have an image of some white text on a blue background. Actually the text could be any color and the background color be any color. The text has some aliasing on it as well. I wish to find the color of the text, so the background color would be one dominant color and the text color is another dominant color. How I can find these two colors?

Really in general I would like to know how to find the two most dominant colors in the image.

I have a image and two points in the image, a and b. How can I use opencv to find a path between the two points which has the least amount of change in pixel value along that path? That is, along that path there is only a minimal change in pixel value from a to b? There are many ways to get from a to b, but some of these ways are going to have more change in pixel value than others.

I used grabCut after the findContours and this seems to work well in about 75% of cases, I have not fine tuned it by altering parameters yet. In most cases where it is not working its due to the findContours and associate code not finding the bounding rectangle correctly.

I'll try this tomorrow and let you know how it goes.

How to use OpenCV (Python script) to remove the background from these button images so that only the button remains? I wish to be able to re-use the button image to place it onto a different background and not sure what OpenCV methods to use. Here are some example button images:

Here is a zip file with many button images to try: http://www.orsa-studio.com/buttons.zip

Note that buttons have the following properties.

1. Their background may not be completely flat with one color, it may be a pattern or a gradient.
2. They have a corner, the corner may have a radius or it may be square.
3. The button will take up most of the image, at least more than 1/2 the width and height of the image.