PiMac's profile - activity

Instead of loading the image you will need to access the one from your camera. You can easily transform IpLImage to Mat

Sorry guys for late answer. I am currently using Common Vision Blox library and using python it is very easy https://for

Hi danielwaldhauser. I didn't do it yet but I will put up one till the end of this month. Please send me your email and

Hi Yusaf. When working with android project you can just import specific version of opencv library to your project. Plea

No problem. I was trying to do sth with

Mat vMat = Mat(Buffers->GetHeight(), Buffers->GetWidth(), CV_8UC(Buffers->GetBytesPerPixel()));

But I don't have a working example so I must have given up when received CVB library. If you figure it out please post the solution if you can.

Hi, please check out OpenCV 3.0 Computer Vision with Java book from packtpub - has low reviews but I believe it contains 80% of knowledge for basic usage of OpenCV in Java. It comes with examples in java in resources (you can download them from the book site). Usually you can just take any c++/python code and use Java opencv doc to verify what is the implementation of the method in Java. At least that how I am managing it.

I suppose it means Mat or iplimage structure. I will try to give you an example but cant promise anything since I don't recall if I have managed it or not.

Changed OpenCV to 3.1 version and canny works again.

visual studio tutorial

Well I thought that too but it is really easy too use if you give it time.

Hi, To be able to use this camera you have to download Sapera LT 8.12 SDK (or whatever version there is now). As this SDK for me was real pain to use I have used CVB libraries that I have got from my camera provider.

Anyway this is what I got from Teledyne support:

void * pBuffer;

pSaperaBuffer->GetAddress(pSaperaBuffer->GetIndex(), &pBuffer);

"pSaperaBuffer" is the SapBuffer that's used in every Sapera ++ demo.

With that you have the starting address of the Sapera Buffer and that should help you "cast" it to an OpenCV compliant buffer.

Should also work when pBuffer is of type UINT8 (if the Nano outputs 8bit) or UINT16.

Hope this helps - for more, please check the various SapMyProcessing.cpp versions in the Sapera demos\classes\vc samples,

I hope it will help you.

Hi. How about you try using visual studio 2015 ? There is a tutorial on how to import opencv libraries to this ide just need to change *.lib names in linker if you will use opencv 3.

Apparently I had to declare as global

    private Mat edges;
    private List<MatOfPoint> contours;
    private Mat hierarch;

and initialize them in biggestContourSegmentation() method.

    edges = new Mat();
    contours = new ArrayList<>();
    hierarch = new Mat();

Declaring them inside this method gave an error.

Thanks but since some of the blobs are false-positives I really would like someone to suggest a method that can be fed a learning set of images. I am not very advanced in such tasks but I will start with kNN classification algorithm and see how it goes.

My original image is just a plastic form with holes in it. You put fruits in it and backlight shines through them. I just cut out the holes and put them onto a black backgroud so it is easier to see the different holes. The placement is static so the holes are at the same places.

Hi, I would like to detect if there is a black blob in a picture below (school project for detecting defects in fruits):

Right now I just detect if there is a circulat shape (houghcircles) and put 1 if there is and 0 if not - it isn't very accurate. I have to detect the black blobs but some of the blobs are false positive. I have been thinking about teaching an algorithm the good and bad pictures to get the best accuracy. I have only used knn for digit detection would it be a good choice to use it in this exercise? Can you suggest anything other to use?

Thank you!

Unfortunately I think it is not the problem. My code is:

public List<MatOfPoint> biggestContourSegmentation(Mat img) {
    edges = new Mat();
    Imgproc.Canny(img, edges, 30.0, 200.0);

I do believe that it uses

//javadoc: Canny(image, edges, threshold1, threshold2)
public static void Canny(Mat image, Mat edges, double threshold1, double threshold2)
{

    Canny_3(image.nativeObj, edges.nativeObj, threshold1, threshold2);

    return;
}

form Imgproc package and the native class that java calls is actually taking long values:

private static native void Canny_3(long image_nativeObj, long edges_nativeObj, double threshold1, double threshold2);

Hi,

I am trying to modify application "Second sight" from the book Android Application Programming with OpenCV 3 first camera (app works fine ). (Android 7.0 OpenCV 3.3 ) I am using

OpenCVLoader.initAsync(OpenCVLoader.OPENCV_VERSION_3_3_0,
            this, mLoaderCallback);

to load the library. Everything works (Mat, imgproc.cvtcolor) except imgproc.Canny (possibly more) When I call

Imgproc.Canny(img, edges, 30, 200);

I get

java.lang.UnsatisfiedLinkError: No implementation found for void org.opencv.imgproc.Imgproc.Canny_3(long, long, double, double) (tried Java_org_opencv_imgproc_Imgproc_Canny_13 and Java_org_opencv_imgproc_Imgproc_Canny_13__JJDD)

Why is it not working? Please help.

Hi, I would suggest you read this post How to use KNearest in java?.

To actually use it you would need a learning set for text something like this THE MNIST DATABASE of handwritten digits.

Last case is to prepare your image as a (this is for digits.png in the 1st link) thresholded 20x20 square with white digit/letter in the middle of the square.

I too struggle sometimes with lack of examples for Java. Usually the easiest is to convert Python examples.

I am learning OpenCV 3.x using java and came across problem using BRIEF detector:

DescriptorExtractor brief = DescriptorExtractor.create(DescriptorExtractor.BRIEF);
    MatOfKeyPoint keypoints = new MatOfKeyPoint();
    Mat descriptors = new Mat();
    brief.compute(gray, keypoints, descriptors);

    Features2d.drawKeypoints(gray, keypoints, copyMat);

This results in OpenCV Error: Bad argument (Specified descriptor extractor type is not supported.)

Is there a mistake in my code or is BRIEF really not supported?

Thanks.