[英]Rotate camera preview to Portrait Android OpenCV Camera
我正在尝试使用 OpenCV 2.4.3.2 创建相机应用程序并进行一些 opencv 处理。 我希望它能够有多个 UI 方向，而不仅仅是横向。
问题是当我将方向更改为纵向时，图像会横向出现。
我知道我可以在进行图像处理之前旋转输入图像（因此只将方向保留为横向），这很好并且有效，但不能解决我的 UI 的其余部分将处于错误方向的问题.
我也尝试使用此代码将相机旋转 90 度，但它似乎不起作用。
mCamera.setDisplayOrientation(90);
它要么没有效果，要么有时只会导致预览变黑
有没有人用 OpenCV 成功地做到了这一点？ 我的课程从 JavaCameraView 扩展而来。
我做了一个改进，即我在 OpenCV 内部旋转了图像，因为它显示在 CameraBridgeViewBase.java 类中。
在交货和并条方法中：
if (canvas != null) {
canvas.drawColor(0, android.graphics.PorterDuff.Mode.CLEAR);
//canvas.drawBitmap(mCacheBitmap, (canvas.getWidth() - mCacheBitmap.getWidth()) / 2, (canvas.getHeight() - mCacheBitmap.getHeight()) / 2, null);
//Change to support portrait view
Matrix matrix = new Matrix();
matrix.preTranslate((canvas.getWidth() - mCacheBitmap.getWidth()) / 2,(canvas.getHeight() - mCacheBitmap.getHeight()) / 2);
if(getResources().getConfiguration().orientation == Configuration.ORIENTATION_PORTRAIT)
matrix.postRotate(90f,(canvas.getWidth()) / 2,(canvas.getHeight()) / 2);
canvas.drawBitmap(mCacheBitmap, matrix, new Paint());
...基本上，这只是像这样旋转输入图像
这更好，但我显然希望它是全屏的。
我在尝试实现 OpenCV 时遇到了同样的问题。 我能够通过对 deliveryAndDrawFrame 方法进行以下更改来修复它。
旋转画布对象
Canvas canvas = getHolder().lockCanvas(); // Rotate canvas to 90 degrees canvas.rotate(90f, canvas.getWidth()/2, canvas.getHeight()/2);在绘制之前调整位图的大小以适合整个画布大小
// Resize Bitmap bitmap = Bitmap.createScaledBitmap(mCacheBitmap, canvas.getHeight(), canvas.getWidth(), true); // Use bitmap instead of mCacheBitmap canvas.drawBitmap(bitmap, new Rect(0,0,bitmap.getWidth(), bitmap.getHeight()), new Rect( (int)((canvas.getWidth() - mScale*bitmap.getWidth()) / 2), (int)((canvas.getHeight() - mScale*bitmap.getHeight()) / 2), (int)((canvas.getWidth() - mScale*bitmap.getWidth()) / 2 + mScale*bitmap.getWidth()), (int）((canvas.getHeight() - mScale*bitmap.getHeight()) / 2 + mScale*bitmap.getHeight() )), null); // Unlock canvas getHolder().unlockCanvasAndPost(canvas);实际上，您可以使宽度或高度与父级（全屏）匹配。
if (canvas != null) {
Bitmap bitmap = Bitmap.createScaledBitmap(mCacheBitmap, canvas.getHeight(), canvas.getWidth(), true);
canvas.rotate(90,0,0);
float scale = canvas.getWidth() / (float)bitmap.getHeight();
float scale2 = canvas.getHeight() / (float)bitmap.getWidth();
if(scale2 > scale){
scale = scale2;
}
if (scale != 0) {
canvas.scale(scale, scale,0,0);
}
canvas.drawBitmap(bitmap, 0, -bitmap.getHeight(), null);
...
此外，您可以使预览尺寸大于屏幕。 只需修改比例。
我修改了 CameraBridgeViewBase.java 如下：
protected Size calculateCameraFrameSize(List<?> supportedSizes, ListItemAccessor accessor, int surfaceWidth, int surfaceHeight) {
int calcWidth = 0;
int calcHeight = 0;
if(surfaceHeight > surfaceWidth){
int temp = surfaceHeight;
surfaceHeight = surfaceWidth;
surfaceWidth = temp;
}
在函数“deliverAndDrawFrame”中：
if (mScale != 0) {
if(canvas.getWidth() > canvas.getHeight()) {
canvas.drawBitmap(mCacheBitmap, new Rect(0,0,mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2),
(int)((canvas.getHeight() - mScale*mCacheBitmap.getHeight()) / 2),
(int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2 + mScale*mCacheBitmap.getWidth()),
(int)((canvas.getHeight() - mScale*mCacheBitmap.getHeight()) / 2 + mScale*mCacheBitmap.getHeight())), null);
} else {
canvas.drawBitmap(mCacheBitmap, rotateMe(canvas, mCacheBitmap), null);
}
其中rotateMe定义如下：
private Matrix rotateMe(Canvas canvas, Bitmap bm) {
// TODO Auto-generated method stub
Matrix mtx=new Matrix();
float scale = (float) canvas.getWidth() / (float) bm.getHeight();
mtx.preTranslate((canvas.getWidth() - bm.getWidth())/2, (canvas.getHeight() - bm.getHeight())/2);
mtx.postRotate(90,canvas.getWidth()/2, canvas.getHeight()/2);
mtx.postScale(scale, scale, canvas.getWidth()/2 , canvas.getHeight()/2 );
return mtx;
}
与横向模式相比，预览 FPS 较慢，因为计算开销较大。
不幸的是 Opencv4Android 不支持人像相机。 但是有一种方法可以克服它。 1）编写您的自定义相机并将其方向设置为纵向。 2）注册它的预览回调。 3）在onPreviewFrame(byte[]data, Camera camera)创建预览字节的Mat ：
Mat mat = new Mat(previewSize.height, previewSize.width, CvType.CV_8UC1);
mat.put(0, 0, data);
Core.transpose(mat, mat);
Core.flip(mat, mat, -1); // rotates Mat to portrait
CvType取决于您的相机使用的预览格式。
附注。 完成后不要忘记释放您创建的所有 Mat 实例。
聚苯乙烯。 最好在单独的线程上管理您的相机，以免在进行某些检测时使 UI 线程过载。
我有同样的问题，我已经弄清楚了！！ 这是我的解决方案：
作为第一部分，在CameraBridgeViewBase.Java两个构造函数中，添加 WindowManager 的初始化：
public CameraBridgeViewBase(Context context, int cameraId) {
super(context);
mCameraIndex = cameraId;
getHolder().addCallback(this);
mMaxWidth = MAX_UNSPECIFIED;
mMaxHeight = MAX_UNSPECIFIED;
windowManager = (WindowManager) context.getSystemService(Context.WINDOW_SERVICE);
}
public CameraBridgeViewBase(Context context, AttributeSet attrs) {
super(context, attrs);
int count = attrs.getAttributeCount();
Log.d(TAG, "Attr count: " + Integer.valueOf(count));
TypedArray styledAttrs = getContext().obtainStyledAttributes(attrs, R.styleable.CameraBridgeViewBase);
if (styledAttrs.getBoolean(R.styleable.CameraBridgeViewBase_show_fps, false))
enableFpsMeter();
mCameraIndex = styledAttrs.getInt(R.styleable.CameraBridgeViewBase_camera_id, -1);
getHolder().addCallback(this);
mMaxWidth = MAX_UNSPECIFIED;
mMaxHeight = MAX_UNSPECIFIED;
windowManager = (WindowManager) context.getSystemService(Context.WINDOW_SERVICE);
styledAttrs.recycle();
}
然后，您需要将函数deliverAndDrawFrame(CvCameraViewFrame frame)替换如下，
protected void deliverAndDrawFrame(CvCameraViewFrame frame) {
Mat modified;
if (mListener != null) {
modified = mListener.onCameraFrame(frame);
} else {
modified = frame.rgba();
}
boolean bmpValid = true;
if (modified != null) {
try {
Utils.matToBitmap(modified, mCacheBitmap);
} catch (Exception e) {
Log.e(TAG, "Mat type: " + modified);
Log.e(TAG, "Bitmap type: " + mCacheBitmap.getWidth() + "*" + mCacheBitmap.getHeight());
Log.e(TAG, "Utils.matToBitmap() throws an exception: " + e.getMessage());
bmpValid = false;
}
}
if (bmpValid && mCacheBitmap != null) {
Canvas canvas = getHolder().lockCanvas();
if (canvas != null) {
canvas.drawColor(0, android.graphics.PorterDuff.Mode.CLEAR);
int rotation = windowManager.getDefaultDisplay().getRotation();
int degrees = 0;
// config degrees as you need
switch (rotation) {
case Surface.ROTATION_0:
degrees = 90;
break;
case Surface.ROTATION_90:
degrees = 0;
break;
case Surface.ROTATION_180:
degrees = 270;
break;
case Surface.ROTATION_270:
degrees = 180;
break;
}
Matrix matrix = new Matrix();
matrix.postRotate(degrees);
Bitmap outputBitmap = Bitmap.createBitmap(mCacheBitmap, 0, 0, mCacheBitmap.getWidth(), mCacheBitmap.getHeight(), matrix, true);
if (outputBitmap.getWidth() <= canvas.getWidth()) {
mScale = getRatio(outputBitmap.getWidth(), outputBitmap.getHeight(), canvas.getWidth(), canvas.getHeight());
} else {
mScale = getRatio(canvas.getWidth(), canvas.getHeight(), outputBitmap.getWidth(), outputBitmap.getHeight());
}
if (mScale != 0) {
canvas.scale(mScale, mScale, 0, 0);
}
Log.d(TAG, "mStretch value: " + mScale);
canvas.drawBitmap(outputBitmap, 0, 0, null);
if (mFpsMeter != null) {
mFpsMeter.measure();
mFpsMeter.draw(canvas, 20, 30);
}
getHolder().unlockCanvasAndPost(canvas);
}
}
}
并额外添加此功能，
private float getRatio(int widthSource, int heightSource, int widthTarget, int heightTarget) {
if (widthTarget <= heightTarget) {
return (float) heightTarget / (float) heightSource;
} else {
return (float) widthTarget / (float) widthSource;
}
}
没关系，如果这个答案对您有用，请标记“已接受”帮助信誉
这里的所有答案都是黑客。 我更喜欢这个解决方案：
更改 JavaCameraView 代码：
mBuffer = new byte[size];
mCamera.setDisplayOrientation(90); //add this
mCamera.addCallbackBuffer(mBuffer);
第二个变化：
// if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.HONEYCOMB) {
// mSurfaceTexture = new SurfaceTexture(MAGIC_TEXTURE_ID);
// mCamera.setPreviewTexture(mSurfaceTexture);
// } else
// mCamera.setPreviewDisplay(null);
mCamera.setPreviewDisplay(getHolder());
与其他答案一样，我已经编写了我的个人版本的deliveryAndDrawFrame （我还通过评论通知了我的代码开始和结束的位置）：
protected void deliverAndDrawFrame(CvCameraViewFrame frame) {
Mat modified;
if (mListener != null) {
modified = mListener.onCameraFrame(frame);
} else {
modified = frame.rgba();
}
boolean bmpValid = true;
if (modified != null) {
try {
Utils.matToBitmap(modified, mCacheBitmap);
} catch(Exception e) {
Log.e(TAG, "Mat type: " + modified);
Log.e(TAG, "Bitmap type: " + mCacheBitmap.getWidth() + "*" + mCacheBitmap.getHeight());
Log.e(TAG, "Utils.matToBitmap() throws an exception: " + e.getMessage());
bmpValid = false;
}
}
if (bmpValid && mCacheBitmap != null) {
Canvas canvas = getHolder().lockCanvas();
if (canvas != null) {
canvas.drawColor(0, android.graphics.PorterDuff.Mode.CLEAR);
if (BuildConfig.DEBUG) {
Log.d(TAG, "mStretch value: " + mScale);
}
// Start of the fix
Matrix matrix = new Matrix();
matrix.preTranslate( ( canvas.getWidth() - mCacheBitmap.getWidth() ) / 2f, ( canvas.getHeight() - mCacheBitmap.getHeight() ) / 2f );
matrix.postRotate( 90f, ( canvas.getWidth()) / 2f, canvas.getHeight() / 2f );
float scale = (float) canvas.getWidth() / (float) mCacheBitmap.getHeight();
matrix.postScale(scale, scale, canvas.getWidth() / 2f , canvas.getHeight() / 2f );
canvas.drawBitmap( mCacheBitmap, matrix, null );
// Back to original OpenCV code
if (mFpsMeter != null) {
mFpsMeter.measure();
mFpsMeter.draw(canvas, 20, 30);
}
getHolder().unlockCanvasAndPost(canvas);
}
}
}
预览现在处于纵向模式，如您所见：
似乎新的 OpenCV CameraBridgeViewBase.java类太高级了，无法对相机预览的布局提供足够的控制。 看看我的示例代码，它基于一些较旧的 OpenCV 示例并使用纯 Android 代码。 要使用传入onPreviewFrame的字节数组， put()其放入 Mat 并从 YUV 转换为 RGB：
mYuv = new Mat(previewHeight + previewHeight/2, previewWidth, CvType.CV_8UC1);
mYuv.put(0, 0, mBuffer);
Imgproc.cvtColor(mYuv, mRgba, Imgproc.COLOR_YUV420sp2RGBA, 4);
您可能可以在互联网上找到旧的 OpenCV4Android 示例，尽管它们是在几个版本之前取出的。 但是，链接的示例代码和上面的代码段应该足以让您入门。
如果您使用的是 openCV 2.4.9 ，请尝试： 1) 在您的代码中复制 opencv 教程混合处理的内容； 2）纠正不匹配的错误（活动名称和可能的布局参考）； 3) 通过添加android:screenOrientation ="landscape"修改您的清单 4) 更正未成年人错误并运行 !!!! bbaamm（现在应该可以正常工作了）
注意：使用此方法，当手机处于纵向位置时，状态栏会出现在右侧。 由于我们正在开发相机项目，我建议您从预览中删除状态栏。
希望能帮助到你 ！！！
您必须考虑以下几点：
因此，为了获得快速和高分辨率的解决方案，我更改了 JavaCameraView.java 和我的 JNI 部分。 在 JavaCameraView.java 中：
...
if (sizes != null) {
/* Select the size that fits surface considering maximum size allowed */
Size frameSize;
if(width > height)
{
frameSize = calculateCameraFrameSize(sizes, new JavaCameraSizeAccessor(), width, height);
}else{
frameSize = calculateCameraFrameSize(sizes, new JavaCameraSizeAccessor(), height, width);
}
...
mCamera.setParameters(params);
params = mCamera.getParameters();
int bufFrameWidth, bufFrameHeight;
bufFrameWidth = params.getPreviewSize().width;
bufFrameHeight = params.getPreviewSize().height;
if(width > height) {
mFrameWidth = params.getPreviewSize().width;
mFrameHeight = params.getPreviewSize().height;
}else{
mFrameWidth = params.getPreviewSize().height;
mFrameHeight = params.getPreviewSize().width;
}
...
mFrameChain = new Mat[2];
mFrameChain[0] = new Mat(bufFrameHeight + (bufFrameHeight/2), bufFrameWidth, CvType.CV_8UC1);
mFrameChain[1] = new Mat(bufFrameHeight + (bufFrameHeight/2), bufFrameWidth, CvType.CV_8UC1);
AllocateCache();
mCameraFrame = new JavaCameraFrame[2];
mCameraFrame[0] = new JavaCameraFrame(mFrameChain[0], bufFrameWidth, bufFrameHeight);
mCameraFrame[1] = new JavaCameraFrame(mFrameChain[1], bufFrameWidth, bufFrameHeight);
通过这些更改，我们确保使用可用于纵向的最高结果（在 calculateCameraFrameSize 中切换高度/宽度）。 我们仍然将横向作为 onPreviewFrame() 的输入处理，但创建了一个位图以纵向绘制 (AllocateCache)。
最后，我们需要为算法提供肖像帧，以便让他检测“站立”对象并将其返回以保存和渲染位图。 因此，对您的活动进行以下修改：
public Mat rot90(Mat matImage, int rotflag){
//1=CW, 2=CCW, 3=180
Mat rotated = new Mat();
if (rotflag == 1){
rotated = matImage.t();
flip(rotated, rotated, 1); //transpose+flip(1)=CW
} else if (rotflag == 2) {
rotated = matImage.t();
flip(rotated, rotated,0); //transpose+flip(0)=CCW
} else if (rotflag ==3){
flip(matImage, rotated,-1); //flip(-1)=180
} else if (rotflag != 0){ //if not 0,1,2,3:
Log.e(TAG, "Unknown rotation flag("+rotflag+")");
}
return rotated;
}
public Mat onCameraFrame(CvCameraViewFrame inputFrame) {
mRgba = rot90(inputFrame.rgba(), 1);
mGray = rot90(inputFrame.gray(), 1);
...
我已经使用 CameraBridgeViewBase 获得了纵向方向，但我不得不在 OpenCV 中更改 JavaCameraView.java :( 想法是下一个：在相机初始化之后，做下一个
setDisplayOrientation(mCamera, 90);
mCamera.setPreviewDisplay(getHolder());
和 setDisplayOrientation 方法
protected void setDisplayOrientation(Camera camera, int angle){
Method downPolymorphic;
try
{
downPolymorphic = camera.getClass().getMethod("setDisplayOrientation", new Class[] { int.class });
if (downPolymorphic != null)
downPolymorphic.invoke(camera, new Object[] { angle });
}
catch (Exception e1)
{
}
}
“jaiprakashgogi”开发人员回答对我有用。 但问题是预览仍仅保存为横向。 这意味着如果我们将预览设置为 imageview，那么它会显示为横向。
上述解决方案可以将预览显示为纵向，但不会永久保存为纵向。
我通过以下方式解决了这个问题。
请在此处查看我的代码...
public String writeToSDFile(byte[] data, int rotation){
byte[] portraitData=null;
if(rotation==90){
Log.i(TAG,"Rotation is : "+rotation);
Bitmap bitmap= BitmapFactory.decodeByteArray(data,0,data.length);
Matrix matrix = new Matrix();
matrix.postRotate(90);
Bitmap rotatedBitmap = Bitmap.createBitmap(bitmap , 0, 0, bitmap.getWidth(), bitmap.getHeight(), matrix, true);
portraitData=bitmapToByte(rotatedBitmap);
}
File dir=getDirectory();
String imageTime=""+System.currentTimeMillis();
String fileName=Constants.FILE_NAME+imageTime+"."+Constants.IMAGE_FORMAT;
File file = new File(dir, fileName);
try {
FileOutputStream f = new FileOutputStream(file);
if(rotation==90){
f.write(portraitData);
}else {
f.write(data);
}
f.close();
} catch (FileNotFoundException e) {
e.printStackTrace();
Log.i(TAG, "******* File not found. Did you" +
" add a WRITE_EXTERNAL_STORAGE permission to the manifest?");
} catch (IOException e) {
e.printStackTrace();
}
Log.i(TAG,"\n\nFile written to "+file);
return fileName;
}
// convert bitmap to Byte Array
public byte[] bitmapToByte(Bitmap bitmap){
ByteArrayOutputStream outputStream=new ByteArrayOutputStream();
bitmap.compress(Bitmap.CompressFormat.JPEG,100,outputStream);
byte[] array=outputStream.toByteArray();
return array;
}
它完全解决了我的问题。
感谢@Kaye Wrobleski 的回答。 我已经扩展它以允许横向和纵向。 这基本上只是一些额外的代码，可以在提供横向方向的默认代码和他的纵向代码之间轻松切换。
在 CameraBridgeViewBase.java 中插入他的代码作为新方法
protected void deliverAndDrawFramePortrait(CvCameraViewFrame frame) {
Mat modified;
if (mListener != null) {
modified = mListener.onCameraFrame(frame);
} else {
modified = frame.rgba();
}
boolean bmpValid = true;
if (modified != null) {
try {
Utils.matToBitmap(modified, mCacheBitmap);
} catch(Exception e) {
Log.e(TAG, "Mat type: " + modified);
Log.e(TAG, "Bitmap type: " + mCacheBitmap.getWidth() + "*" + mCacheBitmap.getHeight());
Log.e(TAG, "Utils.matToBitmap() throws an exception: " + e.getMessage());
bmpValid = false;
}
}
if (bmpValid && mCacheBitmap != null) {
Canvas canvas = getHolder().lockCanvas();
// Rotate canvas to 90 degrees
canvas.rotate(90f, canvas.getWidth()/2, canvas.getHeight()/2);
if (canvas != null) {
canvas.drawColor(0, android.graphics.PorterDuff.Mode.CLEAR);
Log.d(TAG, "mStretch value: " + mScale);
if (mScale != 0) {
// Resize
Bitmap bitmap = Bitmap.createScaledBitmap(mCacheBitmap, canvas.getHeight(), canvas.getWidth(), true);
// Use bitmap instead of mCacheBitmap
canvas.drawBitmap(bitmap, new Rect(0,0,bitmap.getWidth(), bitmap.getHeight()), new Rect(
(int)((canvas.getWidth() - mScale*bitmap.getWidth()) / 2),
(int)((canvas.getHeight() - mScale*bitmap.getHeight()) / 2),
(int)((canvas.getWidth() - mScale*bitmap.getWidth()) / 2 + mScale*bitmap.getWidth()),
(int)((canvas.getHeight() - mScale*bitmap.getHeight()) / 2 + mScale*bitmap.getHeight())), null);
} else {
Bitmap bitmap = Bitmap.createScaledBitmap(mCacheBitmap, canvas.getHeight(), canvas.getWidth(), true);
// Use bitmap instead of mCacheBitmap
canvas.drawBitmap(bitmap, new Rect(0,0,bitmap.getWidth(), bitmap.getHeight()), new Rect(
(int)((canvas.getWidth() - bitmap.getWidth()) / 2),
(int)((canvas.getHeight() - bitmap.getHeight()) / 2),
(int)((canvas.getWidth() - bitmap.getWidth()) / 2 + bitmap.getWidth()),
(int)((canvas.getHeight() - bitmap.getHeight()) / 2 + bitmap.getHeight())), null);
}
if (mFpsMeter != null) {
mFpsMeter.measure();
mFpsMeter.draw(canvas, 20, 30);
}
getHolder().unlockCanvasAndPost(canvas);
}
}
}
然后修改JavaCameraView.java
添加一个新变量来跟踪我们是处于纵向还是横向模式
private boolean portraitMode;
添加两种设置方向模式的方法
public void setLandscapeMode() {
portraitMode = false;
}
public void setPortraitMode() {
portraitMode = true;
}
现在替换 JavaCameraView CameraWorkerClass 中的这些行，run() 方法
if (!mFrameChain[1 - mChainIdx].empty())
deliverAndDrawFrame(mCameraFrame[1 - mChainIdx]);
使用这些行：
if (!mFrameChain[1 - mChainIdx].empty()) {
if (!portraitMode) {
deliverAndDrawFrame(mCameraFrame[1 - mChainIdx]);
} else {
deliverAndDrawFramePortrait(mCameraFrame[1 - mChainIdx]);
}
}
要在方向之间切换，只需在 JavaCameraView 对象上调用 setLandscapeMode() 或 setPortraitMode() 即可。
请注意，反向纵向和反向横向方向仍然是颠倒的。 您需要将它们旋转 180 度以使它们正面朝上，这可以使用 OpenCV 的 warpAffine() 方法轻松完成。 请注意，使用后置摄像头 (LENS_FACING_BACK) 时，纵向模式会将图像上下颠倒。
也许这可以帮助任何人。 使用 opencv343 在 android 9 上测试。 现在这个全屏和 DETECT 脸在纵向和横向模式下。 CameraBridgeViewBase 类的小改动：
private final Matrix matrix = new Matrix();
......改变deliverAndDrawFrame()方法：
protected void deliverAndDrawFrame(CvCameraViewFrame frame) {
Mat modified;
if (mListener != null) {
modified = mListener.onCameraFrame(frame);
} else {
modified = frame.rgba();
}
boolean bmpValid = true;
if (modified != null) {
try {
Utils.matToBitmap(modified, mCacheBitmap);
} catch(Exception e) {
Log.e(TAG, "Mat type: " + modified);
Log.e(TAG, "Bitmap type: " + mCacheBitmap.getWidth() + "*" + mCacheBitmap.getHeight());
Log.e(TAG, "Utils.matToBitmap() throws an exception: " + e.getMessage());
bmpValid = false;
}
}
if (bmpValid && mCacheBitmap != null) {
int currentOrientation = getResources().getConfiguration().orientation;
if (currentOrientation == Configuration.ORIENTATION_LANDSCAPE) {
Canvas canvas = getHolder().lockCanvas();
if (canvas != null) {
canvas.drawColor(0, android.graphics.PorterDuff.Mode.CLEAR);
if (BuildConfig.DEBUG)
Log.d(TAG, "mStretch value: " + mScale);
if (mScale != 0) {
canvas.drawBitmap(mCacheBitmap, new Rect(0,0,mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2),
(int)((canvas.getHeight() - mScale*mCacheBitmap.getHeight()) / 2),
(int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2 + mScale*mCacheBitmap.getWidth()),
(int)((canvas.getHeight() - mScale*mCacheBitmap.getHeight()) / 2 + mScale*mCacheBitmap.getHeight())), null);
} else {
canvas.drawBitmap(mCacheBitmap, new Rect(0,0,mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((canvas.getWidth() - mCacheBitmap.getWidth()) / 2,
(canvas.getHeight() - mCacheBitmap.getHeight()) / 2,
(canvas.getWidth() - mCacheBitmap.getWidth()) / 2 + mCacheBitmap.getWidth(),
(canvas.getHeight() - mCacheBitmap.getHeight()) / 2 + mCacheBitmap.getHeight()), null);
}
if (mFpsMeter != null) {
mFpsMeter.measure();
mFpsMeter.draw(canvas, 20, 30);
}
getHolder().unlockCanvasAndPost(canvas);
}
} else {
Canvas canvas = getHolder().lockCanvas();
if (canvas != null) {
int saveCount = canvas.save();
canvas.setMatrix(matrix);
mScale = Math.max((float) canvas.getHeight() / mCacheBitmap.getWidth(), (float) canvas.getWidth() / mCacheBitmap.getHeight());
canvas.drawColor(0, android.graphics.PorterDuff.Mode.CLEAR);
if (mScale != 0) {
canvas.drawBitmap(mCacheBitmap, new Rect(0,0,mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((int)((canvas.getWidth() - mCacheBitmap.getWidth()) - mCacheBitmap.getWidth())/2,
(int)(canvas.getHeight() - mScale*mCacheBitmap.getHeight() - mScale*mCacheBitmap.getHeight()/2),
(int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2 + mScale*mCacheBitmap.getWidth()),
(int)((canvas.getHeight() - mCacheBitmap.getHeight()) / 2 + mScale*mCacheBitmap.getHeight())), null);
} else {
canvas.drawBitmap(mCacheBitmap, new Rect(0, 0, mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((canvas.getWidth() - mCacheBitmap.getWidth()) / 2,
(canvas.getHeight() - mCacheBitmap.getHeight()) / 2,
(canvas.getWidth() - mCacheBitmap.getWidth()) / 2 + mCacheBitmap.getWidth(),
(canvas.getHeight() - mCacheBitmap.getHeight()) / 2 + mCacheBitmap.getHeight()), null);
}
canvas.restoreToCount(saveCount);
if (mFpsMeter != null) {
mFpsMeter.measure();
mFpsMeter.draw(canvas, 20, 30);
}
getHolder().unlockCanvasAndPost(canvas);
}
}
}
}
并在 MainActivity 中：
public Mat rotateMat(Mat matImage) {
Mat rotated = matImage.t();
Core.flip(rotated, rotated, 1);
return rotated;
}
@Override
public Mat onCameraFrame(CameraBridgeViewBase.CvCameraViewFrame inputFrame) {
MatOfRect faces = new MatOfRect();
int currentOrientation = getResources().getConfiguration().orientation;
if (currentOrientation == Configuration.ORIENTATION_LANDSCAPE) {
mRgba = inputFrame.rgba();
mGray = inputFrame.gray();
int height = mGray.rows();
if (Math.round(height * 0.2) > 0) {
mFaceSize = (int) Math.round(height * 0.2);
}
cascadeClassifier.detectMultiScale(mGray, faces, 1.1, 3, 2,
new Size(mFaceSize, mFaceSize));
Rect[] facesArray = faces.toArray();
for (int i = 0; i < facesArray.length; i++) {
rectangle(mRgba, facesArray[i].tl(), facesArray[i].br(), FACE_RECT_COLOR, 3);
}
} else {
mRgba = inputFrame.rgba();
mGray = rotateMat(inputFrame.gray());
if (mFaceSize == 0) {
int height = mGray.cols();
if (Math.round(height * 0.2) > 0) {
mFaceSize = (int) Math.round(height * 0.2);
}
}
Mat newMat = rotateMat(mRgba);
if(!isBackCameraOn){
flip(newMat, newMat, -1);
flip(mGray, mGray, -1);
}
if (cascadeClassifier != null)
cascadeClassifier.detectMultiScale(mGray, faces, 1.1, 3, 2, new Size(mFaceSize, mFaceSize));
mGray.release();
Rect[] facesArray = faces.toArray();
for (int i = 0; i < facesArray.length; i++) {
rectangle(newMat, facesArray[i].tl(), facesArray[i].br(), FACE_RECT_COLOR, 3);
}
Imgproc.resize(newMat, mRgba, new Size(mRgba.width(), mRgba.height()));
newMat.release();
}
if(!isBackCameraOn){
flip(mRgba, mRgba, 1);
flip(mGray, mGray, 1);
}
return mRgba;
}
另一种解决方案。 我觉得这更好
protected void deliverAndDrawFrame(CvCameraViewFrame frame) {
Mat modified;
if (mListener != null) {
modified = mListener.onCameraFrame(frame);
} else {
modified = frame.rgba();
}
boolean bmpValid = true;
if (modified != null) {
try {
Utils.matToBitmap(modified, mCacheBitmap);
} catch(Exception e) {
Log.e(TAG, "Mat type: " + modified);
Log.e(TAG, "Bitmap type: " + mCacheBitmap.getWidth() + "*" + mCacheBitmap.getHeight());
Log.e(TAG, "Utils.matToBitmap() throws an exception: " + e.getMessage());
bmpValid = false;
}
}
if (bmpValid && mCacheBitmap != null) {
Canvas canvas = getHolder().lockCanvas();
if (canvas != null) {
canvas.drawColor(0, android.graphics.PorterDuff.Mode.CLEAR);
if (BuildConfig.DEBUG)
Log.d(TAG, "mStretch value: " + mScale);
int currentOrientation = getResources().getConfiguration().orientation;
if (currentOrientation == Configuration.ORIENTATION_LANDSCAPE) {
if (mScale != 0) {
canvas.drawBitmap(mCacheBitmap, new Rect(0,0,mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2),
(int)((canvas.getHeight() - mScale*mCacheBitmap.getHeight()) / 2),
(int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2 + mScale*mCacheBitmap.getWidth()),
(int)((canvas.getHeight() - mScale*mCacheBitmap.getHeight()) / 2 + mScale*mCacheBitmap.getHeight())), null);
} else {
canvas.drawBitmap(mCacheBitmap, new Rect(0,0,mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((canvas.getWidth() - mCacheBitmap.getWidth()) / 2,
(canvas.getHeight() - mCacheBitmap.getHeight()) / 2,
(canvas.getWidth() - mCacheBitmap.getWidth()) / 2 + mCacheBitmap.getWidth(),
(canvas.getHeight() - mCacheBitmap.getHeight()) / 2 + mCacheBitmap.getHeight()), null);
}
} else {
if (mScale != 0) {
Bitmap bitmap = Bitmap.createScaledBitmap(mCacheBitmap, canvas.getHeight(), canvas.getWidth(), true);
canvas.drawBitmap(bitmap, new Rect(0,0,bitmap.getWidth(), bitmap.getHeight()), new Rect(
(int)((canvas.getWidth() - mScale*bitmap.getWidth()) / 2),
(int)(0),
(int)((canvas.getWidth() - mScale*bitmap.getWidth()) / 2 + mScale*bitmap.getWidth()),
(int)((canvas.getHeight()))), null);
} else {
Bitmap bitmap = Bitmap.createScaledBitmap(mCacheBitmap, canvas.getHeight(), canvas.getWidth(), true);
canvas.drawBitmap(bitmap, new Rect(0,0,bitmap.getWidth(), bitmap.getHeight()), new Rect(
(int)((canvas.getWidth() - bitmap.getWidth()) / 2),
(int)(0),
(int)((canvas.getWidth() - bitmap.getWidth()) / 2 + bitmap.getWidth()),
(int)(canvas.getHeight())), null);
}
}
if (mFpsMeter != null) {
mFpsMeter.measure();
mFpsMeter.draw(canvas, 20, 30);
}
getHolder().unlockCanvasAndPost(canvas);
}
}
}
和...
@Override
public Mat onCameraFrame(CameraBridgeViewBase.CvCameraViewFrame inputFrame) {
MatOfRect faces = new MatOfRect();
int currentOrientation = getResources().getConfiguration().orientation;
if (currentOrientation == Configuration.ORIENTATION_LANDSCAPE) {
mRgba = inputFrame.rgba();
mGray = inputFrame.gray();
int height = mGray.rows();
if (Math.round(height * 0.2) > 0) {
mFaceSize = (int) Math.round(height * 0.2);
}
cascadeClassifier.detectMultiScale(mGray, faces, 1.1, 3, 2,
new Size(mFaceSize, mFaceSize));
Rect[] facesArray = faces.toArray();
for (int i = 0; i < facesArray.length; i++) {
Point center = new Point(facesArray[i].x + facesArray[i].width / 2,
facesArray[i].y + facesArray[i].height / 2);
rectangle(mRgba, facesArray[i].tl(), facesArray[i].br(), FACE_RECT_COLOR, 3);
}
} else {
mRgba = inputFrame.rgba();
mGray = inputFrame.gray();
Mat rotImage = Imgproc.getRotationMatrix2D(new Point(mRgba.cols() / 2,
mRgba.rows() / 2), 90, 1.0);
Imgproc.warpAffine(mRgba, mRgba, rotImage, mRgba.size());
Imgproc.warpAffine(mGray, mGray, rotImage, mRgba.size());
Core.flip(mRgba, mRgba, 1);
Core.flip(mGray, mGray, 1);
int height = mGray.rows();
if (Math.round(height * 0.2) > 0) {
mFaceSize = (int) Math.round(height * 0.2);
}
cascadeClassifier.detectMultiScale(mGray, faces, 1.1, 3, 2,
new Size(mFaceSize, mFaceSize));
Rect[] facesArray = faces.toArray();
for (int i = 0; i < facesArray.length; i++) {
Point center = new Point(facesArray[i].x + facesArray[i].width / 2,
facesArray[i].y + facesArray[i].height / 2);
rectangle(mRgba, facesArray[i].tl(), facesArray[i].br(), FACE_RECT_COLOR, 3);
}
}
return mRgba;
我不认为有办法做到这一点，没有一些像素操作。 然而，如果我们简单地修改所有这些像素被绘制到的矩阵。 答案部分在于 CameraBridgeViewBase.java 文件
1.进入CameraBridgeViewBase类
2. 制作函数更新矩阵
private final Matrix mMatrix = new Matrix();
private void updateMatrix() {
float mw = this.getWidth();
float mh = this.getHeight();
float hw = this.getWidth() / 2.0f;
float hh = this.getHeight() / 2.0f;
float cw = (float)Resources.getSystem().getDisplayMetrics().widthPixels; //Make sure to import Resources package
float ch = (float)Resources.getSystem().getDisplayMetrics().heightPixels;
float scale = cw / (float)mh;
float scale2 = ch / (float)mw;
if(scale2 > scale){
scale = scale2;
}
boolean isFrontCamera = mCameraIndex == CAMERA_ID_FRONT;
mMatrix.reset();
if (isFrontCamera) {
mMatrix.preScale(-1, 1, hw, hh); //MH - this will mirror the camera
}
mMatrix.preTranslate(hw, hh);
if (isFrontCamera){
mMatrix.preRotate(270);
} else {
mMatrix.preRotate(90);
}
mMatrix.preTranslate(-hw, -hh);
mMatrix.preScale(scale,scale,hw,hh);
}
3. 覆盖 onMeasure 和 layout 功能
@Override
public void layout(int l, int t, int r, int b) {
super.layout(l, t, r, b);
updateMatrix();
}
@Override
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
super.onMeasure(widthMeasureSpec, heightMeasureSpec);
updateMatrix();
}
4.替换现有的deliverAndDrawFrame函数
protected void deliverAndDrawFrame(CvCameraViewFrame frame) { //replaces existing deliverAndDrawFrame
Mat modified;
if (mListener != null) {
modified = mListener.onCameraFrame(frame);
} else {
modified = frame.rgba();
}
boolean bmpValid = true;
if (modified != null) {
try {
Utils.matToBitmap(modified, mCacheBitmap);
} catch(Exception e) {
Log.e(TAG, "Mat type: " + modified);
Log.e(TAG, "Bitmap type: " + mCacheBitmap.getWidth() + "*" + mCacheBitmap.getHeight());
Log.e(TAG, "Utils.matToBitmap() throws an exception: " + e.getMessage());
bmpValid = false;
}
}
if (bmpValid && mCacheBitmap != null) {
Canvas canvas = getHolder().lockCanvas();
if (canvas != null) {
canvas.drawColor(0, android.graphics.PorterDuff.Mode.CLEAR);
int saveCount = canvas.save();
canvas.setMatrix(mMatrix);
if (mScale != 0) {
canvas.drawBitmap(mCacheBitmap, new Rect(0,0,mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2),
(int)((canvas.getHeight() - mScale*mCacheBitmap.getHeight()) / 2),
(int)((canvas.getWidth() - mScale*mCacheBitmap.getWidth()) / 2 + mScale*mCacheBitmap.getWidth()),
(int)((canvas.getHeight() - mScale*mCacheBitmap.getHeight()) / 2 + mScale*mCacheBitmap.getHeight())), null);
} else {
canvas.drawBitmap(mCacheBitmap, new Rect(0,0,mCacheBitmap.getWidth(), mCacheBitmap.getHeight()),
new Rect((canvas.getWidth() - mCacheBitmap.getWidth()) / 2,
(canvas.getHeight() - mCacheBitmap.getHeight()) / 2,
(canvas.getWidth() - mCacheBitmap.getWidth()) / 2 + mCacheBitmap.getWidth(),
(canvas.getHeight() - mCacheBitmap.getHeight()) / 2 + mCacheBitmap.getHeight()), null);
}
//Restore canvas after draw bitmap
canvas.restoreToCount(saveCount);
if (mFpsMeter != null) {
mFpsMeter.measure();
mFpsMeter.draw(canvas, 20, 30);
}
getHolder().unlockCanvasAndPost(canvas);
}
}
}
经过大量的搜索，我发现了这个——
https://answers.opencv.org/question/23972/face-detect-with-portrait-mode-and-front-camera/
这有效。
另外，请确保您在 AndroidManifest.xml 中设置了纵向模式
android:screenOrientation="portrait"
我不太清楚，但相机尺寸是由屏幕宽度决定的。 由于屏幕宽度较低，因此在纵向方向上，相机高度也由低决定。 因此，相机分辨率也由低决定。 并且预览图像放下（预览图像旋转由CameraBridgeViewBase.java中的相机图像的宽度和高度决定）。
作为解决方案，使用横向（将 manifest.xml 中的横向模式确定为 Activity）。 结果，因为屏幕宽度很高，高度也会很高，你的应用程序决定了高分辨率。 此外，您不必旋转相机图像并始终处于全屏模式。 但缺点是原点不同。 我尝试了高分辨率图像作为纵向的多种方法，但我找不到方法。
我的应用程序：纵向
我的相机图像是 720、480/横向 1280、1080。
如本页所述，修改JavaCameraView.java的代码
这真的很容易修复。
Log.d(TAG, "startPreview");
mCamera.startPreview();
Log.d(TAG, "startPreview");
setDisplayOrientation(mCamera, 90);
mCamera.setPreviewDisplay(getHolder());
mCamera.startPreview();
将相机预览旋转到Portrait Android OpenCV Camera和Xamarin
[英]Rotate camera preview to Portrait Android OpenCV Camera and Xamarin
声明:本站的技术帖子网页，遵循CC BY-SA 4.0协议，如果您需要转载，请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.