User1's profile - activity

Can I use CascadeClassifier with SVM also ??

BUT svm.load() should take the path of the training file ?!

I have a problem with load() method in SVM module. For example if I run code:

try {
    Uri uri = Uri.fromFile(new File("file:///android_asset/SVM_Bills.xml"));
    mySvm.load(uri.getPath());          
    int in = mySvm.get_support_vector_count();
    if (in != 0) {
        Toast.makeText(context,"done loading : " + uri.getPath(), Toast.LENGTH_SHORT).show();
        value = (int)mySvm.predict(matOfDescriptor);
    }
    else {
        Toast.makeText(context,"error while loading : " + uri.getPath(), Toast.LENGTH_SHORT).show();
    }

It give me the error toast message (i.e it goes to else part and show the toast message). Of course SVM was previously trained.

Any help please?

I wanna load svm file from assets in order to use it to make prediction. BUT how to do that in android ?

I try these lines in java and they are works well

CvSVM svm = new CvSVM();

svm.load(PathToFile);

I'm working on a feature extraction/matching app using opencv on android . I'm trying to use those lines in my android app but the app crash even if the variables featureDetector & descriptorExtractor are not NULL.

FeatureDetector featureDetector = FeatureDetector.create(FeatureDetector.SIFT);

DescriptorExtractor descriptorExtractor = DescriptorExtractor.create(DescriptorExtractor.SIFT);

it crashes exactly when I call this line

featureDetector.detect(image,matOfKeyPoint);

"Imgproc.resize" function does not work in android application.

I wrote a java test program to run on my PC, and the resize method worked perfectly?

Any help why ? Or how to solve problem ?

YES I increase the min distance as u said :)))

Thank u :)

I follow the instructions here but i still get this error

"Could not find OpenCV Library - 2.4.11.apk!"

When I run the app it tells me that the open cv manger package was not found !!! try to install it !!!

How can I get the number of circles that return by hough circle using java open CV.

Mat circles = new Mat();

Imgproc.HoughCircles(grayImage, circles, Imgproc.CV_HOUGH_GRADIENT, 1 , gray.rows()/2 , 200, 20, 30, 0 );

System.out.println("#rows " + circles.rows() + " #cols " + circles.cols());

I don't know how it store the circles in Mat that has cols and rows ??

I try to get the circle from column but it does not work !!!

if (circles.cols() > 0)
    for (int x = 0; x < circles.cols(); x++) 
    {
        double vCircle[] = circles.get(0,x);
        if (vCircle == null)
            break;
        Point pt = new Point(Math.round(vCircle[0]),
        Math.round(vCircle[1]));
        int radius = (int)Math.round(vCircle[2]);
        // draw the found circle
        Core.circle(imgRGB, pt, radius, new Scalar(0,255,0), 2);
        Core.circle(imgRGB, pt, 3, new Scalar(0,255,0), 2);
     }

For example look at this image

Any help please.

open followed by close give me better result :-) thank u a lot :)

U mean I need to use Filter by size to get rid of all small blobs ?! Could u please tell me the name of filter that can I use for this purpose ?

Yes I try Erode and Dilate But it does not work well ...

I'm working on a java image processing program(based on OpenCV libraryJava). I need to remove the small regions from the image to get a clean image that contain only the large circle ?

i.e Is there a way to get the confidence values in SVM prediction ?

I'm developing an matching system in java using opencv the system accept an image with one object which is coin in order to determine the value of the currency.

What I did : 1-Convert image to gray level. 2-Use edge detection in the gray image the output of this step is used as input to the next step. 3-Use SIFT to find the feature descriptors. 4-Use SVM by storing the SIFT descriptors to predict the value of image.

BUT this algorithm does not give good result. 1-what can I do in order to enhance this algorithm, 2- what is the best SVM params that is suitable for this case . 3-what is the appropriate edge detection for this purpose.

Any suggestion please. Note that in the number of coins types are five.

Start first by image with one coin only then if the matching system that u use is good then u can crop each coin and pass it to your matching system.

In your matching system you can use SIFT to detect key points and find descriptors then use any matcher such as FlANN matcher to find the matches.

Finally u can use RANSAC as criteria if there is good match or not . Hope this help.