Rui Marques's profile - activity

Is there a tutorial for someone who wants to contribute with some Java wrapping native code?

I know this is a bit weird question but i actually came to a situation where i would need this. So I appreciate the patience if someone knows the answer =)

From these parameteres:

int fontFace, 
double fontScale
int fontThickness 
int fontLineType

Is it possible to calculate the width of a character in pixels?

Note 1: If the font is not of fixed width, one can assume it is the letter 'a' or do any other assumption.

Note 2:The idea is not to have some image processing doing this, the idea is just to do some math with the putText() parameters.

In a case were you only want relative areas, which one is faster to compute: calculate a contour area - contourArea() - or count the number of ROI non-zero pixels?

You can assume the contours have been found and you have everything needed to do either case.

Do you have any reason against using JSON in the PC instead of XML?

I do not know if I understood your question correctly, you want to convert XML to JSON? Why would you do that?

ORB is quite slower than FAST (but faster than other detectors like SIFT, SUFT). It is also much more robust than FAST, regarding view-point changes, etc.

Those frame-rates you mention seem normal, you have some degree of tunning, for example getting rid of outliers early so the rest of your code has less keypoints to deal with.

Here are some nice places to put computer vision jobs, I think we should not further fragment the job seeking/posting market otherwise the interested parties have to check 10 's sites:

http://www.computervisiononline.com/jobs http://cvisioncentral.com/jobs/ http://careers.stackoverflow.com

That seems like a problem that would work nicely with simple approaches like applying some morphological operation like dilate and then do watershed to segment the black blobs.

You did the hardest part: having big black blobs which are the drops.

It will add that it would be nice to have better support for network cameras and mobile video encoding. Some kind of SLAM module (I think some work is being done here), or state of the art object detection/tracking.

In the context of matching features from a live camera with features from an image of an object, how do you filter the detected features and/or the matches?

I prefer answers for OpenCV4Android, but answers for the desktop version are also acceptable!

So it is not open source?

That is more clear, I do not know why you didn't ask that way in the first place :p

You can do something like this:

//  List<MatOfPoint> contours; from findCountours  

for(int i = 0; i<contours.size(); i++) {
    MatOfPoint m = contours.get(i);
    List<Point> list = m.toList();
    for(int j = 0; j<list.size(); j++) {
        Point p = list.get(j);
        if(p.x == 500){

        }
    }
}

That site is deprecated AFAIK. The documentation is now at: http://docs.opencv.org

You can check the answers for this post:

http://answers.opencv.org/question/8873/best-way-to-store-a-mat-object-in-android/

I recommend the answer I gave there, with some small changes it can be applied to other OpenCV data structures besides Mat.

If you really do not want to convert your data to json you can check there how to convert to a byte array. After you have a byte array you can store it in a file or whatever.

This is the best answer for Java, very complete.

When I use the FAST detector on an image of the size 960 x 720, I get about 9000 points with the current method. I was wondering if there is a technique to set a limit on the number of points that are being detected? Like the 500 best points only? I know that reducing the resolution gives less points, but I want to know if there is another way with a higher resolution?

FeatureDetector fast = FeatureDetector.create(FeatureDetector.FAST);

fast.detect(mIntermediateMat, points);

// gives +9000 points

This would help me in the description phase to reduce the calculation time to describe all these points. (takes about 0.8 seconds at the moment, and the descriptor is bigger then the original image when looking at the Mb). I don't use the native part, but the Java code and the OpenCV Library 2.4.4.

EDIT: I found this article on how to describe ORB parameters, and was wondering if it might work for FAST.

I added

FeatureDetector fast = FeatureDetector.create(FeatureDetector.FAST);

    String filename = "mnt/sdcard/fast_params.yml";
                    WriteFile(filename, "%YAML:1.0\nimage: mIntermediateMat\nkeypoints: points\nthreshold : 300 \nnonmaxSupression : true\ntype : FastFeatureDetector::TYPE_9_16\n");
                    DescriptorExtractor extractor = null;
                    extractor.read(filename);
    fast.detect(mIntermediateMat, points);

The string is based upon information i found here but it is giving back a NullPointerException and I'm not use it even works for FAST. Anyone?

Yes sorry, I was assuming a buffer filled with the pixels only.