Develop Reference

libav sws_scale() fails colorspace conversion on real device, works on emulator

I'm making a movie player with libav. I have decoding video packets working, I have play in reverse working, I have seeking working. All this works no an x86 android emulator, but fails to work on a real android phone (arm64-v8a)
The failure is in sws_scale() - it returns 0. The video frames continue to be decoded properly with no errors.
There are no errors, warnings, alerts from libav. I have connected an avlog_callback
void log_callback(void *ptr, int level, const char *fmt, va_list vargs) {
if (level<= AV_LOG_WARNING)
__android_log_print( level, LOG_TAG, fmt, vargs);
}
uint64_t openMovie( char* path, int rotate, float javaDuration )
{
av_log_set_level(AV_LOG_WARNING);
av_log_set_callback(log_callback);
The code to do the sws_scale() is:
int JVM_getBitmapBuffer( JNIEnv* env, jobject thiz, jlong av, jobject bufferAsInt, jbyte transparent ) {
avblock *block = (avblock *) av;
if (!block) {
__android_log_print(ANDROID_LOG_ERROR, LOG_TAG, " avblock is null");
return AVERROR(EINVAL);
}
if (!block->pCodecCtx) {
__android_log_print(ANDROID_LOG_ERROR, LOG_TAG, " codecctx is null");
return AVERROR(EINVAL);
}
int width = block->pCodecCtx->width;
int height = block->pCodecCtx->height;
if (NULL == block->sws) {
__android_log_print( ANDROID_LOG_ERROR, LOG_TAG, "getBitmapBuffer:\n *** invalid sws context ***" );
}
int scaleRet = sws_scale( block->sws,
block->pFrame->data,
block->pFrame->linesize,
0,
height,
block->pFrameRGB->data,
block->pFrameRGB->linesize
);
if (scaleRet == 0 ) {
__android_log_print(ANDROID_LOG_ERROR, LOG_TAG, " scale failed");
__android_log_print(ANDROID_LOG_ERROR, LOG_TAG, " pframe linesize %d", block->pFrame->linesize[0]);
__android_log_print(ANDROID_LOG_ERROR, LOG_TAG, " pframergb linesize %d", block->pFrameRGB->linesize[0]);
__android_log_print(ANDROID_LOG_ERROR, LOG_TAG, " height %d",
height);
return AVERROR(EINVAL);
}
Setting up the codex and avframes:
//i have tried every combination of 1, 8, 16, and 32 for these values
int alignRGB = 32;
int align = 16;
int width = block->pCodecCtx->width;
int height = block->pCodecCtx->height;
block->pFrame = av_frame_alloc();
block->pFrameRGB = av_frame_alloc();
block->pFrameRGBBuffer = av_malloc(
(size_t)av_image_get_buffer_size(AV_PIX_FMT_RGB32, width, height, alignRGB)
);
av_image_fill_arrays(
block->pFrameRGB->data,
block->pFrameRGB->linesize,
block->pFrameRGBBuffer,
AV_PIX_FMT_RGB32,
width,
height,
alignRGB
);
block->pFrameBuffer = av_malloc(
(size_t) av_image_get_buffer_size(block->pCodecCtx->pix_fmt,
width, height, align
)
);
av_image_fill_arrays(
block->pFrame->data,
block->pFrame->linesize,
block->pFrameBuffer,
block->pCodecCtx->pix_fmt,
width, height,
align
);
block->sws = sws_getContext(
width, height,
AV_PIX_FMT_YUV420P,
width, height,
AV_PIX_FMT_RGB32,
SWS_BILINEAR, NULL, NULL, 0
);
Wildcards are that:
I'm using React-Native
My emulator is x86 android api 28
My real-device is arm64-v8a AOSP (around api 28, don't remember exactly(
Other notes:
libav .so files are compiled from mobile-ffmpeg project.
I can also sws_scale also works on x86_64 linux using SDL to project YV12
Test video is here: https://github.com/markkimsal/video-thumbnailer/tree/master/fixtures
block is a simple C struct with pointers to relevant AV memory structures.
Using FFMPEG 4.3.2
I'm pretty certain it has something to do with the pixel alignment. But documentation is practically non-existent on this topic. It could also be the difference between pixel formats RGBA and RGB32, or possibly little-endian vs big-endian.

This is a known bug in FFMPEG on ARM architecture.
A workaround was posted by mythtv that involves subtracting 1 from the destination width in order to bypass broken optimization code.
https://code.mythtv.org/trac/ticket/12888
https://code.mythtv.org/trac/changeset/7de03a90c1b144fc0067261af1c9cfdd8d358972/mythtv
Reported against FFMPEG 3.2.1
http://trac.ffmpeg.org/ticket/6192
Still exists in FFMPEG 4.3.2
int new_width = width;
#if ARCH_ARM
// The ARM build of FFMPEG has a bug that if sws_scale is
// called with source and dest sizes the same, and
// formats as shown below, it causes a bus error and the
// application core dumps. To avoid this I make a -1
// difference in the new width, causing it to bypass
// the code optimization which is failing.
if (pix_fmt == AV_PIX_FMT_YUV420P
&& dst_pix_fmt == AV_PIX_FMT_BGRA)
new_width = width - 1;
#endif
d->swsctx = sws_getCachedContext(d->swsctx, width, height, pix_fmt,
new_width, height, dst_pix_fmt,
SWS_FAST_BILINEAR, NULL, NULL, NULL);
UPDATE:
Building mobile-ffmpeg 4.3.2 with --debug option produces a working binary.
UPDATE 2:
The --debug flag does not seem to be working anymore. Must've been some unclean builds. This does seem to be related to alpha channel.
The only non-alpha RGB format supported by both Android.Bitmap.Config and ffmpeg is RGB_565. Using 565 works with even numbered destination widths.
in libswscale/yuv2rgb.c:
SwsFunc ff_yuv2rgb_get_ptr(SwsContext *c) {
...
switch(c->dstFormat) {
case AV_PIX_FMT_RGBA:
return (CONFIG_SWSCALE_ALPHA & isALPHA(c->srcFormat)) ? yuva2rgb_c: yuv2rgb_c_32;
....
case AV_PIX_FMT_RGB565:
return yuv2rgb_c_16_ordered_dither;
...
My source is YUV420P, and does not support alpha channels (like yuv A 420p). When my destination format is RGB565, this method seems to not have any problems.
It's a shame that I can't produce any alpha channel for Android. I guess the real answer is to just use OpenGL and use a surface that supports blitting YUV directly.
UPDATE 3
I have found the codepaths in libswscale/swscale_unscaled.c and libswscale/yuv2rgb.c
If you scale the output then you can convert to RGBA dest format.
If you add the flag SWS_ACCURATE_RND to your SwsContext then you also avoid the bad codepath.

Rendering issue on Project Tango using OpenCV image processing

I came across one problem to render the camera image after some process on its YUV buffer.
I am using the example video-overlay-jni-example and in the method OnFrameAvailable I am creating a new frame buffer using the cv::Mat...
Here is how I create a new frame buffer:
cv::Mat frame((int) yuv_height_ + (int) (yuv_height_ / 2), (int) yuv_width_, CV_8UC1, (uchar *) yuv_temp_buffer_.data());
After process, I copy the frame.data to the yuv_temp_buffer_ in order to render it on the texture: memcpy(&yuv_temp_buffer_[0], frame.data, yuv_size_);
And this works fine...
The problem starts when I try to execute an OpenCV method findChessboardCorners... using the frame that I've created before.
The method findChessboardCorners takes about 90ms to execute (11 fps), however, it seems to be rendering in a slower rate. (It appears to be rendering in ~0.5 fps on the screen).
Here is the code of the OnFrameAvailable method:
void AugmentedRealityApp::OnFrameAvailable(const TangoImageBuffer* buffer) {
if (yuv_drawable_ == NULL){
return;
}
if (yuv_drawable_->GetTextureId() == 0) {
LOGE("AugmentedRealityApp::yuv texture id not valid");
return;
}
if (buffer->format != TANGO_HAL_PIXEL_FORMAT_YCrCb_420_SP) {
LOGE("AugmentedRealityApp::yuv texture format is not supported by this app");
return;
}
// The memory needs to be allocated after we get the first frame because we
// need to know the size of the image.
if (!is_yuv_texture_available_) {
yuv_width_ = buffer->width;
yuv_height_ = buffer->height;
uv_buffer_offset_ = yuv_width_ * yuv_height_;
yuv_size_ = yuv_width_ * yuv_height_ + yuv_width_ * yuv_height_ / 2;
// Reserve and resize the buffer size for RGB and YUV data.
yuv_buffer_.resize(yuv_size_);
yuv_temp_buffer_.resize(yuv_size_);
rgb_buffer_.resize(yuv_width_ * yuv_height_ * 3);
AllocateTexture(yuv_drawable_->GetTextureId(), yuv_width_, yuv_height_);
is_yuv_texture_available_ = true;
}
std::lock_guard<std::mutex> lock(yuv_buffer_mutex_);
memcpy(&yuv_temp_buffer_[0], buffer->data, yuv_size_);
///
cv::Mat frame((int) yuv_height_ + (int) (yuv_height_ / 2), (int) yuv_width_, CV_8UC1, (uchar *) yuv_temp_buffer_.data());
if (!stam.isCalibrated()) {
Profiler profiler;
profiler.startSampling();
stam.initFromChessboard(frame, cv::Size(9, 6), 100);
profiler.endSampling();
profiler.print("initFromChessboard", -1);
}
///
memcpy(&yuv_temp_buffer_[0], frame.data, yuv_size_);
swap_buffer_signal_ = true;
}
Here is the code of the method initFromChessBoard:
bool STAM::initFromChessboard(const cv::Mat& image, const cv::Size& chessBoardSize, int squareSize)
{
cv::Mat rvec = cv::Mat(cv::Size(3, 1), CV_64F);
cv::Mat tvec = cv::Mat(cv::Size(3, 1), CV_64F);
std::vector<cv::Point2d> imagePoints, imageBoardPoints;
std::vector<cv::Point3d> boardPoints;
for (int i = 0; i < chessBoardSize.height; i++)
{
for (int j = 0; j < chessBoardSize.width; j++)
{
boardPoints.push_back(cv::Point3d(j*squareSize, i*squareSize, 0.0));
}
}
//getting only the Y channel (many of the functions like face detect and align only needs the grayscale image)
cv::Mat gray(image.rows, image.cols, CV_8UC1);
gray.data = image.data;
bool found = findChessboardCorners(gray, chessBoardSize, imagePoints, cv::CALIB_CB_FAST_CHECK);
#ifdef WINDOWS_VS
printf("Number of chessboard points: %d\n", imagePoints.size());
#elif ANDROID
LOGE("Number of chessboard points: %d", imagePoints.size());
#endif
for (int i = 0; i < imagePoints.size(); i++) {
cv::circle(image, imagePoints[i], 6, cv::Scalar(149, 43, 0), -1);
}
}
Is anyone having the same problem after process something in the YUV buffer to render on the texture?
I did a test using other device rather than the project Tango using camera2 API, and the rendering process on the screen appears to be the same rate of the OpenCV function process itself.
I appreciate any help.

I had a similar problem. My app slowed down after using the copied yuv buffer and doing some image processing with OpenCV. I would recommand you to use the tango_support library to access the yuv image buffer by doing the following:
In your config function:
int AugmentedRealityApp::TangoSetupConfig() {
TangoSupport_createImageBufferManager(TANGO_HAL_PIXEL_FORMAT_YCrCb_420_SP, 1280, 720, &yuv_manager_);
}
In your callback function:
void AugmentedRealityApp::OnFrameAvailable(const TangoImageBuffer* buffer) {
TangoSupport_updateImageBuffer(yuv_manager_, buffer);
}
In your render thread:
void AugmentedRealityApp::Render() {
TangoImageBuffer* yuv = new TangoImageBuffer();
TangoSupport_getLatestImageBuffer(yuv_manager_, &yuv);
cv::Mat yuv_frame, rgb_img, gray_img;
yuv_frame.create(720*3/2, 1280, CV_8UC1);
memcpy(yuv_frame.data, yuv->data, 720*3/2*1280); // yuv image
cv::cvtColor(yuv_frame, rgb_img, CV_YUV2RGB_NV21); // rgb image
cvtColor(rgb_img, gray_img, CV_RGB2GRAY); // gray image
}
You can share the yuv_manger with other objects/threads so you can access the yuv image buffer wherever you want.

GrayScaled Image losing its brightness in NDK(C/C++) Android

I am passing the Bitmap with ARGB_8888 config.
I am able to apply the grayscale effect to the image but after applying that i am losing its brightness.
I have googled a lot but found the same implementation as i have.
Here is my native implmentation ::
JNIEXPORT void JNICALL Java_com_example_ndksampleproject_MainActivity_jniConvertToGray(JNIEnv * env, jobject obj, jobject bitmapcolor,jobject bitmapgray)
{
AndroidBitmapInfo infocolor;
void* pixelscolor;
AndroidBitmapInfo infogray;
void* pixelsgray;
int ret;
int y;
int x;
LOGI("convertToGray");
if ((ret = AndroidBitmap_getInfo(env, bitmapcolor, &infocolor)) < 0) {
LOGE("AndroidBitmap_getInfo() failed ! error=%d", ret);
return;
}
if ((ret = AndroidBitmap_getInfo(env, bitmapgray, &infogray)) < 0) {
LOGE("AndroidBitmap_getInfo() failed ! error=%d", ret);
return;
}
LOGI("color image :: width is %d; height is %d; stride is %d; format is %d;flags is %d",infocolor.width,infocolor.height,infocolor.stride,infocolor.format,infocolor.flags);
if (infocolor.format != ANDROID_BITMAP_FORMAT_RGBA_8888) {
LOGE("Bitmap format is not RGBA_8888 !");
return;
}
LOGI("gray image :: width is %d; height is %d; stride is %d; format is %d;flags is %d",infogray.width,infogray.height,infogray.stride,infogray.format,infogray.flags);
if (infogray.format != ANDROID_BITMAP_FORMAT_A_8) {
LOGE("Bitmap format is not A_8 !");
return;
}
if ((ret = AndroidBitmap_lockPixels(env, bitmapcolor, &pixelscolor)) < 0) {
LOGE("AndroidBitmap_lockPixels() failed ! error=%d", ret);
}
if ((ret = AndroidBitmap_lockPixels(env, bitmapgray, &pixelsgray)) < 0) {
LOGE("AndroidBitmap_lockPixels() failed ! error=%d", ret);
}
LOGI("unlocking pixels height = %d",infocolor.height);
// modify pixels with image processing algorithm
for (y=0;y<infocolor.height;y++) {
argb * line = (argb *) pixelscolor;
uint8_t * grayline = (uint8_t *) pixelsgray;
for (x=0;x<infocolor.width;x++) {
grayline[x] = ((255-0.3 * line[x].red) + (255-0.59 * line[x].green) + (255-0.11*line[x].blue))/3;
}
pixelscolor = (char *)pixelscolor + infocolor.stride;
pixelsgray = (char *) pixelsgray + infogray.stride;
}
LOGI("unlocking pixels");
AndroidBitmap_unlockPixels(env, bitmapcolor);
AndroidBitmap_unlockPixels(env, bitmapgray);
}
Result ::
Please let me know if you need anything from my side..
Please help me to get rid of this issue as i am stuck into this from many hours.
Many thanks in Advance !!!
EDIT ::
After applying the Mark Setchell's Suggestion ::
EDITED
If you invert the image above, you get this - which looks correct to me:

Don't divide by 3 on the line where you calculate grayline[x]. Your answer is already correctly weighted because 0.3 + 0.59 + 0.11 = 1
grayline[x] = (255-0.3 * line[x].red) + (255-0.59 * line[x].green) + (255-0.11*line[x].blue);

There are two problems with your current code.
1) As mentioned by others, do not divide the final result by three. If you were calculating grayscale using the average method (e.g. gray = (R + G + B) / 3), the division would be necessary. For the ITU conversion formula you are using, there is no need for this extra division, because the fractional amounts already sum to 1.
2) The inversion occurs because you are subtracting each color value from 255. There is no need to do this.
The correct grayscale conversion code for your current formula would be:
grayline[x] = ((0.3 * line[x].red) + (0.59 * line[x].green) + (0.11*line[x].blue));

Load .png image as ARGB_8888 bitmap [duplicate]

I try to read an image from sdcard (in emulator) and then create a Bitmap image with the
BitmapFactory.decodeByteArray
method. I set the options:
options.inPrefferedConfig = Bitmap.Config.ARGB_8888
options.inDither = false
Then I extract the pixels into a ByteBuffer.
ByteBuffer buffer = ByteBuffer.allocateDirect(width*height*4)
bitmap.copyPixelsToBuffer(buffer)
I use this ByteBuffer then in the JNI to convert it into RGB format and want to calculate on it.
But always I get false data - I test without modifying the ByteBuffer. Only thing I do is to put it into the native method into JNI. Then cast it into a unsigned char* and convert it back into a ByteBuffer before returning it back to Java.
unsigned char* buffer = (unsinged char*)(env->GetDirectBufferAddress(byteBuffer))
jobject returnByteBuffer = env->NewDirectByteBuffer(buffer, length)
Before displaying the image I get data back with
bitmap.copyPixelsFromBuffer( buffer )
But then it has wrong data in it.
My Question is if this is because the image is internally converted into RGB 565 or what is wrong here?
.....
Have an answer for it:
->>> yes, it is converted internally to RGB565.
Does anybody know how to create such an bitmap image from PNG with ARGB8888 pixel format?
If anybody has an idea, it would be great!

An ARGB_8888 Bitmap (on pre Honeycomb versions) is natively stored in the RGBA format.
So the alpha channel is moved at the end. You should take this into account when accessing a Bitmap's pixels natively.
I assume you are writing code for a version of Android lower than 3.2 (API level < 12), because since then the behavior of the methods
BitmapFactory.decodeFile(pathToImage);
BitmapFactory.decodeFile(pathToImage, opt);
bitmapObject.createScaledBitmap(bitmap, desiredWidth, desiredHeight, false /*filter?*/);
has changed.
On older platforms (API level < 12) the BitmapFactory.decodeFile(..) methods try to return a Bitmap with RGB_565 config by default, if they can't find any alpha, which lowers the quality of an iamge. This is still ok, because you can enforce an ARGB_8888 bitmap using
options.inPrefferedConfig = Bitmap.Config.ARGB_8888
options.inDither = false
The real problem comes when each pixel of your image has an alpha value of 255 (i.e. completely opaque). In that case the Bitmap's flag 'hasAlpha' is set to false, even though your Bitmap has ARGB_8888 config. If your *.png-file had at least one real transparent pixel, this flag would have been set to true and you wouldn't have to worry about anything.
So when you want to create a scaled Bitmap using
bitmapObject.createScaledBitmap(bitmap, desiredWidth, desiredHeight, false /*filter?*/);
the method checks whether the 'hasAlpha' flag is set to true or false, and in your case it is set to false, which results in obtaining a scaled Bitmap, which was automatically converted to the RGB_565 format.
Therefore on API level >= 12 there is a public method called
public void setHasAlpha (boolean hasAlpha);
which would have solved this issue. So far this was just an explanation of the problem.
I did some research and noticed that the setHasAlpha method has existed for a long time and it's public, but has been hidden (#hide annotation). Here is how it is defined on Android 2.3:
/**
* Tell the bitmap if all of the pixels are known to be opaque (false)
* or if some of the pixels may contain non-opaque alpha values (true).
* Note, for some configs (e.g. RGB_565) this call is ignore, since it does
* not support per-pixel alpha values.
*
* This is meant as a drawing hint, as in some cases a bitmap that is known
* to be opaque can take a faster drawing case than one that may have
* non-opaque per-pixel alpha values.
*
* #hide
*/
public void setHasAlpha(boolean hasAlpha) {
nativeSetHasAlpha(mNativeBitmap, hasAlpha);
}
Now here is my solution proposal. It does not involve any copying of bitmap data:
Checked at runtime using java.lang.Reflect if the current
Bitmap implementation has a public 'setHasAplha' method.
(According to my tests it works perfectly since API level 3, and i haven't tested lower versions, because JNI wouldn't work). You may have problems if a manufacturer has explicitly made it private, protected or deleted it.
Call the 'setHasAlpha' method for a given Bitmap object using JNI.
This works perfectly, even for private methods or fields. It is official that JNI does not check whether you are violating the access control rules or not.
Source: http://java.sun.com/docs/books/jni/html/pitfalls.html (10.9)
This gives us great power, which should be used wisely. I wouldn't try modifying a final field, even if it would work (just to give an example). And please note this is just a workaround...
Here is my implementation of all necessary methods:
JAVA PART:
// NOTE: this cannot be used in switch statements
private static final boolean SETHASALPHA_EXISTS = setHasAlphaExists();
private static boolean setHasAlphaExists() {
// get all puplic Methods of the class Bitmap
java.lang.reflect.Method[] methods = Bitmap.class.getMethods();
// search for a method called 'setHasAlpha'
for(int i=0; i<methods.length; i++) {
if(methods[i].getName().contains("setHasAlpha")) {
Log.i(TAG, "method setHasAlpha was found");
return true;
}
}
Log.i(TAG, "couldn't find method setHasAlpha");
return false;
}
private static void setHasAlpha(Bitmap bitmap, boolean value) {
if(bitmap.hasAlpha() == value) {
Log.i(TAG, "bitmap.hasAlpha() == value -> do nothing");
return;
}
if(!SETHASALPHA_EXISTS) { // if we can't find it then API level MUST be lower than 12
// couldn't find the setHasAlpha-method
// <-- provide alternative here...
return;
}
// using android.os.Build.VERSION.SDK to support API level 3 and above
// use android.os.Build.VERSION.SDK_INT to support API level 4 and above
if(Integer.valueOf(android.os.Build.VERSION.SDK) <= 11) {
Log.i(TAG, "BEFORE: bitmap.hasAlpha() == " + bitmap.hasAlpha());
Log.i(TAG, "trying to set hasAplha to true");
int result = setHasAlphaNative(bitmap, value);
Log.i(TAG, "AFTER: bitmap.hasAlpha() == " + bitmap.hasAlpha());
if(result == -1) {
Log.e(TAG, "Unable to access bitmap."); // usually due to a bug in the own code
return;
}
} else { //API level >= 12
bitmap.setHasAlpha(true);
}
}
/**
* Decodes a Bitmap from the SD card
* and scales it if necessary
*/
public Bitmap decodeBitmapFromFile(String pathToImage, int pixels_limit) {
Bitmap bitmap;
Options opt = new Options();
opt.inDither = false; //important
opt.inPreferredConfig = Bitmap.Config.ARGB_8888;
bitmap = BitmapFactory.decodeFile(pathToImage, opt);
if(bitmap == null) {
Log.e(TAG, "unable to decode bitmap");
return null;
}
setHasAlpha(bitmap, true); // if necessary
int numOfPixels = bitmap.getWidth() * bitmap.getHeight();
if(numOfPixels > pixels_limit) { //image needs to be scaled down
// ensures that the scaled image uses the maximum of the pixel_limit while keeping the original aspect ratio
// i use: private static final int pixels_limit = 1280*960; //1,3 Megapixel
imageScaleFactor = Math.sqrt((double) pixels_limit / (double) numOfPixels);
Bitmap scaledBitmap = Bitmap.createScaledBitmap(bitmap,
(int) (imageScaleFactor * bitmap.getWidth()), (int) (imageScaleFactor * bitmap.getHeight()), false);
bitmap.recycle();
bitmap = scaledBitmap;
Log.i(TAG, "scaled bitmap config: " + bitmap.getConfig().toString());
Log.i(TAG, "pixels_limit = " + pixels_limit);
Log.i(TAG, "scaled_numOfpixels = " + scaledBitmap.getWidth()*scaledBitmap.getHeight());
setHasAlpha(bitmap, true); // if necessary
}
return bitmap;
}
Load your lib and declare the native method:
static {
System.loadLibrary("bitmaputils");
}
private static native int setHasAlphaNative(Bitmap bitmap, boolean value);
Native section ('jni' folder)
Android.mk:
LOCAL_PATH := $(call my-dir)
include $(CLEAR_VARS)
LOCAL_MODULE := bitmaputils
LOCAL_SRC_FILES := bitmap_utils.c
LOCAL_LDLIBS := -llog -ljnigraphics -lz -ldl -lgcc
include $(BUILD_SHARED_LIBRARY)
bitmapUtils.c:
#include <jni.h>
#include <android/bitmap.h>
#include <android/log.h>
#define LOG_TAG "BitmapTest"
#define Log_i(...) __android_log_print(ANDROID_LOG_INFO,LOG_TAG,__VA_ARGS__)
#define Log_e(...) __android_log_print(ANDROID_LOG_ERROR,LOG_TAG,__VA_ARGS__)
// caching class and method IDs for a faster subsequent access
static jclass bitmap_class = 0;
static jmethodID setHasAlphaMethodID = 0;
jint Java_com_example_bitmaptest_MainActivity_setHasAlphaNative(JNIEnv * env, jclass clazz, jobject bitmap, jboolean value) {
AndroidBitmapInfo info;
void* pixels;
if (AndroidBitmap_getInfo(env, bitmap, &info) < 0) {
Log_e("Failed to get Bitmap info");
return -1;
}
if (info.format != ANDROID_BITMAP_FORMAT_RGBA_8888) {
Log_e("Incompatible Bitmap format");
return -1;
}
if (AndroidBitmap_lockPixels(env, bitmap, &pixels) < 0) {
Log_e("Failed to lock the pixels of the Bitmap");
return -1;
}
// get class
if(bitmap_class == NULL) { //initializing jclass
// NOTE: The class Bitmap exists since API level 1, so it just must be found.
bitmap_class = (*env)->GetObjectClass(env, bitmap);
if(bitmap_class == NULL) {
Log_e("bitmap_class == NULL");
return -2;
}
}
// get methodID
if(setHasAlphaMethodID == NULL) { //initializing jmethodID
// NOTE: If this fails, because the method could not be found the App will crash.
// But we only call this part of the code if the method was found using java.lang.Reflect
setHasAlphaMethodID = (*env)->GetMethodID(env, bitmap_class, "setHasAlpha", "(Z)V");
if(setHasAlphaMethodID == NULL) {
Log_e("methodID == NULL");
return -2;
}
}
// call java instance method
(*env)->CallVoidMethod(env, bitmap, setHasAlphaMethodID, value);
// if an exception was thrown we could handle it here
if ((*env)->ExceptionOccurred(env)) {
(*env)->ExceptionDescribe(env);
(*env)->ExceptionClear(env);
Log_e("calling setHasAlpha threw an exception");
return -2;
}
if(AndroidBitmap_unlockPixels(env, bitmap) < 0) {
Log_e("Failed to unlock the pixels of the Bitmap");
return -1;
}
return 0; // success
}
That's it. We are done. I've posted the whole code for copy-and-paste purposes.
The actual code isn't that big, but making all these paranoid error checks makes it a lot bigger. I hope this could be helpful to anyone.

How to cast an IplImage* to a cv:Mat*?

I´m a begginer on C Language and I need to copy pixels to Android Bitmap, I´m using a piece of code of android opencv, used for a jni:
AndroidBitmapInfo info;
void* pixels;
int ret;
cv::Mat* mat;
if ((ret = AndroidBitmap_getInfo(env, bitmap, &info)) < 0 ){
LOGE("AndroidBitmap_getInfo() failed ! error=%d", ret);
return false; // can't get info
}
if (info.format != ANDROID_BITMAP_FORMAT_RGBA_8888){
LOGE("Bitmap format is not RGB_8888 !");
return false; // incompatible format
}
if ( (ret = AndroidBitmap_lockPixels(env, bitmap, &pixels)) < 0 ){
LOGE("AndroidBitmap_lockPixels() failed ! error=%d", ret);
return false; // can't get pixels
}
memcpy(pixels, mat->data, info.height * info.width * 4);
AndroidBitmap_unlockPixels(env, bitmap);
So, I have an IplImage* called pImage, but I don´t know how to convert an IplImage* to a cv::Mat*. I see a way to convert to a cv:Mat, like this:
cv::Mat mat(pImage);
But I need an cv:Mat* not an cv:Mat. Any help?

Answering to your question title:
cv::Mat* mat = new cv::Mat(pImage);

The OpenCV tutorials include an article on Interoperability with OpenCV 1.