来自OpenCV的摄像头的Andr​​oid模板匹配 [英] OpenCV Android template matching from camera

查看:235
本文介绍了来自OpenCV的摄像头的Andr​​oid模板匹配的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

我想要的图像匹配在Android上的摄像机输入。当我尝试用2图像一切工作就好了。但现在我喜欢做同样的事情用相机输入。要完成这件事,我实现CvCameraViewListener2并尝试以下code:

I'm trying to match an image with the camera input in Android. When I try this with 2 images everything works just fine. But now I like to do the same thing with the camera input. To get this done I implemented CvCameraViewListener2 and tried the following code:

@Override
public Mat onCameraFrame(CvCameraViewFrame inputFrame) {
    mRgba = inputFrame.rgba();
    int match_method = Imgproc.TM_CCOEFF;

    mSizeRgba = mRgba.size();

    int rows = (int) mSizeRgba.height;
    int cols = (int) mSizeRgba.width;

    Mat templ = Highgui.imread(getFileAbsPath("template.jpg"));

    // Create the result matrix
    int result_cols = cols - templ.cols() + 1;
    int result_rows = rows - templ.rows() + 1;

    Mat result = new Mat(result_rows, result_cols, CvType.CV_32F);

    Mat src = new Mat(result_rows, result_cols, CvType.CV_32F);
    mRgba.convertTo(src, CvType.CV_32F);

    Mat template = new Mat(templ.rows(), templ.cols(), CvType.CV_32F);
    templ.convertTo(template, CvType.CV_32F);

    // Do the Matching and Normalize
    Imgproc.matchTemplate(src, templ, result, match_method);
    Core.normalize(result, result, 0, 1, Core.NORM_MINMAX, -1, new Mat());

    // Localizing the best match with minMaxLoc
    MinMaxLocResult mmr = Core.minMaxLoc(result);

    Point matchLoc;
    if (match_method == Imgproc.TM_SQDIFF
            || match_method == Imgproc.TM_SQDIFF_NORMED) {
        matchLoc = mmr.minLoc;
    } else {
        matchLoc = mmr.maxLoc;
    }
    Rect roi = new Rect((int) matchLoc.x, (int) matchLoc.y, templ.cols(),
            templ.rows());
    // Mat cropped = new Mat(mRgba, roi);
    Core.rectangle(result, new Point(roi.x, roi.y), new Point(roi.width - 2, roi.height - 2), new Scalar(255, 0, 0, 255), 2);
    return mRgba;
}

当我运行此code我得到这个OpenCV的错误:

When I run this code I get this OpenCV error:

OpenCV Error: Assertion failed ((img.depth() == CV_8U || img.depth() == CV_32F) 
                                          && img.type() == templ.type()) in ...

谁能帮我解决这个问题?

Can anyone help me with this problem?

感谢

推荐答案

我解决我的问题。我需要模板的颜色转换BGR为RGBA。 用下面的$ C $词有没有崩溃了,但在preVIEW相机帧速度非常慢。这不正是我想要的。

I solved my problem. I needed to convert the color of the template from BGR to RGBA. With the following code i have no crashes anymore but the camera frames in the preview are very slow. This is not exactly what I want.

    public void initialize(){
    if (src.empty())
        return;
    if(template == null){
        Mat templ = Highgui.imread(getFileAbsPath("template.png"), Highgui.CV_LOAD_IMAGE_UNCHANGED);
        template = new Mat(templ.size(), CvType.CV_32F);
        Imgproc.cvtColor(templ, template, Imgproc.COLOR_BGR2RGBA);
    }
}

@Override
public Mat onCameraFrame(CvCameraViewFrame inputFrame) {

    src = inputFrame.rgba();
    initialize();
    int match_method = Imgproc.TM_SQDIFF;

    // Create the result matrix
    int result_cols = src.cols() - template.cols() + 1;
    int result_rows = src.rows() - template.rows() + 1;
    Mat result = new Mat(result_rows, result_cols, CvType.CV_32F);

    // Do the Matching and Normalize
    Imgproc.matchTemplate(src, template, result, match_method);
    Core.normalize(result, result, 0, 1, Core.NORM_MINMAX, -1, new Mat());

    // Localizing the best match with minMaxLoc
    MinMaxLocResult mmr = Core.minMaxLoc(result);

    Point matchLoc;
    if (match_method == Imgproc.TM_SQDIFF || match_method == Imgproc.TM_SQDIFF_NORMED) {
        matchLoc = mmr.minLoc;
    } else {
        matchLoc = mmr.maxLoc;
    }

    Rect roi = new Rect((int) matchLoc.x, (int) matchLoc.y, template.cols(), template.rows());
    Core.rectangle(src, new Point(roi.x, roi.y), new Point(roi.width - 2, roi.height - 2), new Scalar(255, 0, 0, 255), 2);
    return src;
}

这篇关于来自OpenCV的摄像头的Andr​​oid模板匹配的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆