Develop Reference

Taking Picture via intent to internal (app) storage returns null for bitmap

The scope: I want to take a picture via intent and save the picture to the internal storage of my app.
Then I want to load a scaled version into a byte array (from inputstream), save this scaled image as byte array into SQLight.
After saving it to the database I want to delete the picture.
(This question only is about saving the image to internal storage, the scope is only here because there is always someone that ask about it)
The problem: I'm stuck at saving the picture to the internal storage.
I'll add examples from my debugging session as comments behind the variables to show the values i got while testing.
I have an ImageView which has an onClickListener that starts the takePictureIntent:
With following global attributes:
Uri mCurrentPhotoUri; //URI to file
File mCurrentPicture; //the current picture don't know if I need it somewhere but for complete understanding of code
imageView.setOnClickListener(new View.OnClickListener() {
#Override
public void onClick(View v) {
//Intent for the on-board camera
Intent takePictureIntent = new Intent(MediaStore.ACTION_IMAGE_CAPTURE);
//device has camera
if(takePictureIntent.resolveActivity(getPackageManager()) != null) {
File photoFile = null;
try {
//create a file with path the code below
photoFile = createImageFile(); //sets photoFile to: /data/data/my.app.project/app_photo/JPEG_20151105_092219_-1434131481.jpg
} catch (IOException e) {
e.printStackTrace();
}
//file has been created, set members and add Extra to intent, then start intent.
if(photoFile != null) {
mCurrentPicture = photoFile; // well, same as above
mCurrentPhotoUri = Uri.fromFile(photoFile); // this looks somehow wrong, but I don't know much about URIs: file:///data/data/my.app.project/app_photo/JPEG_20151105_092219_-1434131481.jpg
takePictureIntent.putExtra(MediaStore.EXTRA_OUTPUT, Uri.fromFile(photoFile)); //same URI as above that extra should be needed to tell the cam that I don't want to save to the default path but my app path
startActivityForResult(takePictureIntent, 10); //start the intent and use requestcode 10 for onActivityResult ...
}
}
}
});
The creation of the file path:
//code from google developers with some changes.
private File createImageFile() throws IOException {
String timestamp = new SimpleDateFormat("yyyyMMdd_HHmmss").format(new Date()); //from today value: 20151105_092219
String imageFilename = "JPEG_" + timestamp + "_"; // concat is this: JPEG_20151105_092219_
File storageDir = this.getDir("photo", MODE_PRIVATE); //String path is: /data/data/my.app.project/app_photo
storageDir.mkdirs();
File image = File.createTempFile(imageFilename, ".jpg", storageDir); //String path is: /data/data/my.app.project/app_photo/JPEG_20151105_092219_-1434131481.jpg
mCurrentPhotoPath = "file:" + image.getAbsolutePath(); //here I put the absolute path into static mCurrentPhotoPath, concate with the "file:" from googledeveloper guide: file:/data/data/my.app.project/app_photo/JPEG_20151105_092219_-1434131481.jpg
return image;
}
So the camera opens and I can take a picture and I'm ask if I want to save that picture (all from the build-in camera app, device is a samsung galaxy note).
Then my onActivityResult-Method is called:
I used data as parameter because I used the mini byte array for something, but with the custom storage this returns null and it isn't used anymore.
#Override
protected void onActivityResult(int requestCode, int resultCode, Intent data) {
if (resultCode == RESULT_OK) {
switch(requestcode) {
...
case 10:
setImageView(ivPreview1, data, 0);
ivPreview.setVisibility(View.VISIBLE);
break;
...
}
...
}
}
Method setImageView:
private void setImageView(ImageView iv, Intent data, int index) {
try {
Uri u = mCurrentPhotoUri; //sets u to: file:///data/data/my.app.project/app_photo/JPEG_20151105_092219_-1434131481.jpg
File file = new File(u.getPath()); //sets file to: /data/data/my.app.project/app_photo/JPEG_20151105_092219_-1434131481.jpg
Bitmap bm = null;
ByteArrayOutputStream baos = null;
int orientation = 0;
if (file.exists()) { //this is true
//found that somewhere in the developer training:
ExifInterface exif = null;
try {
exif = new ExifInterface(photoUri.getPath());
} catch (IOException e) {
e.printStackTrace();
}
if(exif != null)
orientation = exif.getAttributeInt(ExifInterface.TAG_ORIENTATION, 0); //is 0 (i didn't rotate the tablet)
//resulution I want to resize the image to:
int reqWidth = 960, reqHeight = 1280;
//exchange values if orientation doesn't match landscape
if (orientation == 0 || orientation == 270) {
int temp = reqWidth;
reqWidth = reqHeight;
reqHeight = temp;
}
//this I used before I changed to internal storage to change the size of the image code below
bm = ImageManager.decodeSampledBitmapFromFile(u.getPath(), reqWidth, reqHeight); // returns null because of this everything following is null too.
if (orientation == 90 || orientation == 180 || orientation == 270) {
Matrix matrix = new Matrix();
// rotate the Bitmap
if (orientation == 90)
matrix.postRotate(90F);
else if (orientation == 270)
matrix.postRotate(-90F);
else
matrix.postRotate(180F);
// recreate the new Bitmap
bm = Bitmap.createBitmap(bm, 0, 0,
bm.getWidth(), bm.getHeight(), matrix, true);
}
baos = new ByteArrayOutputStream();
bm.compress(Bitmap.CompressFormat.JPEG, 50, baos);
}
iv.setImageBitmap(bm);
} catch (Exception e) {
e.printStackTrace();
Log.e(TAG, "Could not take Photo: ", e);
}
}
The following methods that I used to decode the file (customisation of: http://developer.android.com/downloads/samples/DisplayingBitmaps.zip ):
The line with BitmapFactory.decodeFile(filename, options); also creates a log entry: D/skia: --- SkImageDecoder::Factory returned null
public static Bitmap decodeSampledBitmapFromFile(String filename,
int reqWidth, int reqHeight) {
//this gets parameters:
// reqHeight: 960, reqWidth: 1280 and filename: /data/data/my.app.project/app_photo/JPEG_20151105_092219_-1434131481.jpg
// First decode with inJustDecodeBounds=true to check dimensions
final BitmapFactory.Options options = new BitmapFactory.Options();
options.inJustDecodeBounds = true;
BitmapFactory.decodeFile(filename, options); // this adds outHeight and outWidth to -1 (variables from options)
//this also creates a log entry: D/skia: --- SkImageDecoder::Factory returned null
// Calculate inSampleSize
options.inSampleSize = calculateInSampleSize(options, reqWidth, reqHeight);
// Decode bitmap with inSampleSize set
options.inJustDecodeBounds = false;
Bitmap bmp = BitmapFactory.decodeFile(filename, options);
return bmp;
}
public static int calculateInSampleSize(BitmapFactory.Options options,
int reqWidth, int reqHeight) {
// BEGIN_INCLUDE (calculate_sample_size)
// Raw height and width of image
final int height = options.outHeight; //is -1
final int width = options.outWidth; //is -1
int inSampleSize = 1;
//because its obviously smaller in both statements code will not be executed so it returns 1
if (height > reqHeight || width > reqWidth) {
final int halfHeight = height / 2;
final int halfWidth = width / 2;
// Calculate the largest inSampleSize value that is a power of 2 and keeps both
// height and width larger than the requested height and width.
while ((halfHeight / inSampleSize) > reqHeight
&& (halfWidth / inSampleSize) > reqWidth) {
inSampleSize *= 2;
}
// This offers some additional logic in case the image has a strange
// aspect ratio. For example, a panorama may have a much larger
// width than height. In these cases the total pixels might still
// end up being too large to fit comfortably in memory, so we should
// be more aggressive with sample down the image (=larger inSampleSize).
long totalPixels = width * height / inSampleSize;
// Anything more than 2x the requested pixels we'll sample down further
final long totalReqPixelsCap = reqWidth * reqHeight * 2;
while (totalPixels > totalReqPixelsCap) {
inSampleSize *= 2;
totalPixels /= 2;
}
}
return inSampleSize;
// END_INCLUDE (calculate_sample_size)
}
I'm stuck at this for several days now I don't have any ideas that could solve my problem. This is also due to lack of android knowledge and the fact that i can't use emulators on my pc so i can't even look in the app folder to see if a picture was taken.

Try to get the path for storing temporary image like below.which will return your app folder location.and add the permission as well.
File dir = context.getExternalFilesDir(null)+"/"+"photo";
Add uses-feature for camera access too.
<manifest ... >
<uses-feature android:name="android.hardware.camera"
android:required="true" />
Official documentation.

How to rotate a JPEG file on Android without losing quality and gaining file size?

Background
I need to rotate images taken by the camera so that they will always have a normal orientation.
for this, I use the next code (used this post to get the image orientation)
//<= get the angle of the image , and decode the image from the file
final Matrix matrix = new Matrix();
//<= prepare the matrix based on the EXIF data (based on https://gist.github.com/9re/1990019 )
final Bitmap rotatedBitmap = Bitmap.createBitmap(bitmap, 0, 0, bitmap.getWidth(), bitmap.getHeight(),matrix,false);
bitmap.recycle();
fileOutputStream = new FileOutputStream(tempFilePath);
rotatedBitmap.compress(CompressFormat.JPEG, 100, fileOutputStream);
rotatedBitmap.recycle();
here the compression rate (AKA "quality" parameter) is 100.
The problem
The code works fine, but the result is larger than the original, much much larger.
The original file is around 600-700 KB, while the resulting file is around 3MB ...
This is even though both the input file and the output file are of the same format (JPEG).
The camera settings are at "super fine" quality. not sure what it means, but I think it has something to do with the compression ratio.
What I've tried
I've tried to set the "filter" parameter to either false or true. both resulted in large files.
Even without the rotation itself (just decode and encode), I get much larger files sizes...
Only when I've set compression ratio to around 85, I get similar files sizes, but I wonder how the quality is affected compared to the original files.
The question
Why does it occur?
Is there a way to get the exact same size and quality of the input file ?
Will using the same compression rate as the original file make it happen? Is it even possible to get the compression rate of the original file?
What does it mean to have a 100% compression rate ?
EDIT: I've found this link talking about rotation of JPEG files without losing the quality and file size , but is there a solution for it on Android ?
Here's another link that says it's possible, but I couldn't find any library that allows rotation of jpeg files without losing their quality

I tried two methods but I found out those methods take too long in my case. I still share what I used.
Method 1: LLJTran for Android
Get the LLJTran from here:
https://github.com/bkhall/AndroidMediaUtil
The code:
public static boolean rotateJpegFileBaseOnExifWithLLJTran(File imageFile, File outFile){
try {
int operation = 0;
int degree = getExifRotateDegree(imageFile.getAbsolutePath());
//int degree = 90;
switch(degree){
case 90:operation = LLJTran.ROT_90;break;
case 180:operation = LLJTran.ROT_180;break;
case 270:operation = LLJTran.ROT_270;break;
}
if (operation == 0){
Log.d(TAG, "Image orientation is already correct");
return false;
}
OutputStream output = null;
LLJTran llj = null;
try {
// Transform image
llj = new LLJTran(imageFile);
llj.read(LLJTran.READ_ALL, false); //don't know why setting second param to true will throw exception...
llj.transform(operation, LLJTran.OPT_DEFAULTS
| LLJTran.OPT_XFORM_ORIENTATION);
// write out file
output = new BufferedOutputStream(new FileOutputStream(outFile));
llj.save(output, LLJTran.OPT_WRITE_ALL);
return true;
} catch(Exception e){
e.printStackTrace();
return false;
}finally {
if(output != null)output.close();
if(llj != null)llj.freeMemory();
}
} catch (Exception e) {
// Unable to rotate image based on EXIF data
e.printStackTrace();
return false;
}
}
public static int getExifRotateDegree(String imagePath){
try {
ExifInterface exif;
exif = new ExifInterface(imagePath);
String orientstring = exif.getAttribute(ExifInterface.TAG_ORIENTATION);
int orientation = orientstring != null ? Integer.parseInt(orientstring) : ExifInterface.ORIENTATION_NORMAL;
if(orientation == ExifInterface.ORIENTATION_ROTATE_90)
return 90;
if(orientation == ExifInterface.ORIENTATION_ROTATE_180)
return 180;
if(orientation == ExifInterface.ORIENTATION_ROTATE_270)
return 270;
} catch (IOException e) {
e.printStackTrace();
}
return 0;
}
Method 2: Using libjepg-turbo's jpegtran executable
1 Follow the step describe here:
https://stackoverflow.com/a/12296343/1099884
Except that you don't need obj/local/armeabi/libjpeg.a on ndk-build because I only want the jpegtran executable but not mess with JNI with libjepg.a .
2 Place the jpegtran executable on asset folder.
The code:
public static boolean rotateJpegFileBaseOnExifWithJpegTran(Context context, File imageFile, File outFile){
try {
int operation = 0;
int degree = getExifRotateDegree(imageFile.getAbsolutePath());
//int degree = 90;
String exe = prepareJpegTranExe(context);
//chmod ,otherwise premission denied
boolean ret = runCommand("chmod 777 "+exe);
if(ret == false){
Log.d(TAG, "chmod jpegTran failed");
return false;
}
//rotate the jpeg with jpegtran
ret = runCommand(exe+
" -rotate "+degree+" -outfile "+outFile.getAbsolutePath()+" "+imageFile.getAbsolutePath());
return ret;
} catch (Exception e) {
// Unable to rotate image based on EXIF data
e.printStackTrace();
return false;
}
}
public static String prepareJpegTranExe(Context context){
File exeDir = context.getDir("JpegTran", 0);
File exe = new File(exeDir, "jpegtran");
if(!exe.exists()){
try {
InputStream is = context.getAssets().open("jpegtran");
FileOutputStream os = new FileOutputStream(exe);
int bufferSize = 16384;
byte[] buffer = new byte[bufferSize];
int count;
while ((count=is.read(buffer, 0, bufferSize))!=-1) {
os.write(buffer, 0, count);
}
is.close();
os.close();
} catch (IOException e) {
e.printStackTrace();
}
}
return exe.getAbsolutePath();
}
public static boolean runCommand(String cmd){
try{
Process process = Runtime.getRuntime().exec(cmd);
BufferedReader reader = new BufferedReader(
new InputStreamReader(process.getInputStream()));
int read;
char[] buffer = new char[4096];
StringBuffer output = new StringBuffer();
while ((read = reader.read(buffer)) > 0) {
output.append(buffer, 0, read);
}
reader.close();
// Waits for the command to finish.
process.waitFor();
return true;
} catch (Exception e) {
e.printStackTrace();
return false;
}
}
Unfortunately, both take too long. It is 16 seconds on my Samsung Galaxy S1!!!! But I found out this app (https://play.google.com/store/apps/details?id=com.lunohod.jpegtool) only take 3-4 seconds. There must be some way to do.

Once you are done setting you bestPreviewSize You have to now set for bestPictureSize every phone supports different picture sizes so to get Best Picture quality you have to check supported picture sizes and then set best size to camera parameter. You have to set those parameters in surface changed to get the width and height. surfaceChanged will be called in start and thus your new parameters will be set.
#Override
public void surfaceChanged(SurfaceHolder holder, int format, int width, int height) {
Camera.Parameters myParameters = camera.getParameters();
myPicSize = getBestPictureSize(width, height);
if (myBestSize != null && myPicSize != null) {
myParameters.setPictureSize(myPicSize.width, myPicSize.height);
myParameters.setJpegQuality(100);
camera.setParameters(myParameters);
Toast.makeText(getApplicationContext(),
"CHANGED:Best PICTURE SIZE:\n" +
String.valueOf(myPicSize.width) + " ::: " + String.valueOf(myPicSize.height),
Toast.LENGTH_LONG).show();
}
}
Now the getBestPictureSize ..
private Camera.Size getBestPictureSize(int width, int height)
{
Camera.Size result=null;
Camera.Parameters p = camera.getParameters();
for (Camera.Size size : p.getSupportedPictureSizes()) {
if (size.width>width || size.height>height) {
if (result==null) {
result=size;
} else {
int resultArea=result.width*result.height;
int newArea=size.width*size.height;
if (newArea>resultArea) {
result=size;
}
}
}
}
return result;
}

For rotation, try this..
final Matrix matrix = new Matrix();
matrix.setRotate(90):
final Bitmap rotatedBitmap = Bitmap.createBitmap(bitmap, 0, 0, bitmap.getWidth(), bitmap.getHeight(),matrix,false);
I am using PNG images and it is working fine.... for JPEG images, please check the above code.

100% quality rate probably is a higher quality setting than the setting the files are originally saved with. This results in higher size but (almost) the same image.
I'm not sure how to get exactly the same size, maybe just setting the quality to 85% will do (Quick and Dirty).
However if you just want to rotate the pic in 90°-steps, you could edit just the JPEG-metadata without touching the pixel data itself.
Not sure how it's done in android, but this is how it works.

Decoded bitmap results in larger size

I am trying to store a bitmap (that i have previously read from a file) after decoding it in a preferable size using BitmapFactory.Options.inSampleSize. The problem is that the file size of the stored bitmap is at least double the size of the original file. I have searched a lot and could not find how i can deal with this, since i don't want it to happen for memmory efficiency (later i reuse the stored bitmap). Here is my method that does what i describe:
private Bitmap decodeFileToPreferredSize(File f) {
Bitmap b = null;
try {
// Decode image size
Log.i("Bitmap", "Imported image size: " + f.length() + " bytes");
BitmapFactory.Options o = new BitmapFactory.Options();
o.inJustDecodeBounds = true;
BitmapFactory.decodeFile(f.getAbsolutePath(), o);
//Check if the user has defined custom image size
int scale = 1;
if(pref_height != -1 && pref_width != -1) {
if (o.outHeight > pref_height || o.outWidth > pref_width) {
scale = (int) Math.pow(
2,
(int) Math.round(Math.log(pref_width
/ (double) Math.max(o.outHeight, o.outWidth))
/ Math.log(0.5)));
}
}
// Decode with inSampleSize
BitmapFactory.Options o2 = new BitmapFactory.Options();
o2.inSampleSize = scale;
b = BitmapFactory.decodeFile(f.getAbsolutePath(), o2);
String name = "Image_" + System.currentTimeMillis() + ".jpg";
File file = new File(Environment.getExternalStorageDirectory(), name);
FileOutputStream out = new FileOutputStream(file);
b.compress(Bitmap.CompressFormat.JPEG, 100, out);
out.close();
Log.i("Bitmap", "Exported image size: " + file.length() + " bytes");
} catch (Exception e) {
b = null;
}
return b;
}
UPDATE I saw the densities on the two image files. The one that came from camera intent has 72 dpi density for width and height. The image file created from my above method has 96 dpi density for width and height. This explains why a 0.5 MByte image that came form the camera is resized in approximatelly 2.5 MByte with my above method since the rescale factor is (96/72) * 2 ~= 2.5. For some reason, the bitmap i create does not take the density of the image that came from the camera. I tried to set the density with all variation of BitmapFactory.Options.inDensity but no luck. Also i tried to change the bitmap density with bitmap.setDensity(int dpi); but still no effect. So, my new question is if there is a way to define the density of the bitmap when the image is stored.
Thanks in advance.

I had a similar issue. When I downloaded images from web they used more space on the SD than they did when downloaded to my PC from browser. I think the issue is simply that BitmapFactory saves the images in a non optimzed format of some sort.
My workaround was to use following instead of the bitmapfactory:
try {
try {
is = yourinputstream;
// Consider reading stream twice and return the bitmap.
os = youroutputstream;
byte data[] = new byte[4096];
int count;
while ((count = is.read(data)) != -1) {
os.write(data, 0, count);
}
} catch (Exception e) {
e.printStackTrace();
} finally {
if(is != null) {
is.close();
}
if(os != null) {
os.close();
}
}
} catch (IOException e) {
e.printStackTrace();
}

You are correct that the problem is in a density change.
This thread revealed a solution: BitmapFactory returns bigger image than source
Before decoding also disable inScaled:
options.inSampleSize = 2; //or however you desire, power of 2
options.inScaled = false;
This will keep the density from changing and you will see the decrease in size you were looking for.

When BitmapFactory.Options.inSampleSize = 0.5, it does 100% / 0.5 = 200%.
What you want I think is BitmapFactory.Options.inSampleSize = 2, it does 100% / 2 = 50%

Catching OutOfMemoryError in decoding Bitmap

Is it a good practice to catch OutOfMemoryError even you have tried some ways to reduce memory usage? Or should we just not catching the exception? Which one is better practice?
try {
BitmapFactory.Options options = new BitmapFactory.Options();
options.inSampleSize = 4;
bitmap = BitmapFactory.decodeFile(file, options);
} catch (OutOfMemoryError e) {
e.printStackTrace();
}
Thanks

It's good practice to catch it once and give decodeFile another chance. Catch it and call System.gc() and try decoding again. There is a high probability that it will work after calling System.gc().
try {
BitmapFactory.Options options = new BitmapFactory.Options();
options.inSampleSize = 4;
bitmap = BitmapFactory.decodeFile(file, options);
} catch (OutOfMemoryError e) {
e.printStackTrace();
System.gc();
try {
bitmap = BitmapFactory.decodeFile(file);
} catch (OutOfMemoryError e2) {
e2.printStackTrace();
// handle gracefully.
}
}

I did something like this: I catch the error only for try to scale down the image until it works. Eventually it can not work at all; then returns null; otherwise, in success, returns the bitmap.
Outside I decide what to do with the bitmap whether it's null or not.
// Let w and h the width and height of the ImageView where we will place the Bitmap. Then:
// Get the dimensions of the original bitmap
BitmapFactory.Options bmOptions= new BitmapFactory.Options();
bmOptions.inJustDecodeBounds= true;
BitmapFactory.decodeFile(path, bmOptions);
int photoW= bmOptions.outWidth;
int photoH= bmOptions.outHeight;
// Determine how much to scale down the image.
int scaleFactor= (int) Math.max(1.0, Math.min((double) photoW / (double)w, (double)photoH / (double)h)); //1, 2, 3, 4, 5, 6, ...
scaleFactor= (int) Math.pow(2.0, Math.floor(Math.log((double) scaleFactor) / Math.log(2.0))); //1, 2, 4, 8, ...
// Decode the image file into a Bitmap sized to fill the View
bmOptions.inJustDecodeBounds= false;
bmOptions.inSampleSize= scaleFactor;
bmOptions.inPurgeable= true;
do
{
try
{
Log.d("tag", "scaleFactor: " + scaleFactor);
scaleFactor*= 2;
bitmap= BitmapFactory.decodeFile(path, bmOptions);
}
catch(OutOfMemoryError e)
{
bmOptions.inSampleSize= scaleFactor;
Log.d("tag", "OutOfMemoryError: " + e.toString());
}
}
while(bitmap == null && scaleFactor <= 256);
if(bitmap == null)
return null;
For example, with an image of 3264x2448, the loop iterates 2 times on my phone, and then it works.

You'd want to catch it if you want to display either a smaller image / different image / show a custom error message to the user.
Your image access wrapper can catch these errors and return some custom error codes defined within your code; your activity that uses this code can decide what to do with the error code - warn user, force him to exit with a better error message than the one the android system would provide, etc.
Btw, you are not using the options variable in your sample code.

Though it might not be a good idea to catch OutOfMemoryError using try-catch. But, sometimes you have no choice, because all of us hate app crashes.
So, what you can do is
Catch OutOfMemoryError using try-catch
Since, after this error your activity may become unstable, restart it.
You may disable animations so that user doesn't know that activity is restarted.
You may put some extra data in intent to know that app was crashed during previous run.
How I did is:
try {
//code that causes OutOfMemoryError
} catch (Exception e) {
// in case of exception handle it
e.printStackTrace();
} catch (OutOfMemoryError oome)
{
//restart this activity
Intent i=this.getIntent();
i.addFlags(Intent.FLAG_ACTIVITY_NO_ANIMATION); //disable animation
//EXTRA_ABNORMAL_SHUTDOWN is user defined
i.putExtra(this.EXTRA_ABNORMAL_SHUTDOWN, true);
//put extra data into intent if you like
finish(); //and finish the activity
overridePendingTransition(0, 0);
startActivity(i); //then start it(there is also restart method in newer API)
return false;
}
And then on onCreate of Activity you can resume(something like this):
boolean abnormalShutdown=getIntent().getBooleanExtra(this.EXTRA_ABNORMAL_SHUTDOWN, false);
if (abnormalShutdown)
{
//Alert user for any error
//get any extra data you put befor restarting.
}
This approach saved my app.
Hope it helps you too!!

How do I scale a streaming bitmap in-place without reading the whole image first?

I have an Android application that is very image intensive. I'm currently using Bitmap.createScaledBitmap() to scale the image to a desired size. However, this method requires that I already have the original bitmap in memory, which can be quite sizable.
How can I scale a bitmap that I'm downloading without first writing the entire thing out to local memory or file system?

This method will read the header information from the image to determine its size, then read the image and scale it to the desired size in place without allocating memory for the full original sized image.
It also uses BitmapFactory.Options.inPurgeable, which seems to be a sparsely documented but desirable option to prevent OoM exceptions when using lots of bitmaps. UPDATE: no longer uses inPurgeable, see this note from Romain
It works by using a BufferedInputStream to read the header information for the image before reading the entire image in via the InputStream.
/**
* Read the image from the stream and create a bitmap scaled to the desired
* size. Resulting bitmap will be at least as large as the
* desired minimum specified dimensions and will keep the image proportions
* correct during scaling.
*/
protected Bitmap createScaledBitmapFromStream( InputStream s, int minimumDesiredBitmapWith, int minimumDesiredBitmapHeight ) {
final BufferedInputStream is = new BufferedInputStream(s, 32*1024);
try {
final Options decodeBitmapOptions = new Options();
// For further memory savings, you may want to consider using this option
// decodeBitmapOptions.inPreferredConfig = Config.RGB_565; // Uses 2-bytes instead of default 4 per pixel
if( minimumDesiredBitmapWidth >0 && minimumDesiredBitmapHeight >0 ) {
final Options decodeBoundsOptions = new Options();
decodeBoundsOptions.inJustDecodeBounds = true;
is.mark(32*1024); // 32k is probably overkill, but 8k is insufficient for some jpgs
BitmapFactory.decodeStream(is,null,decodeBoundsOptions);
is.reset();
final int originalWidth = decodeBoundsOptions.outWidth;
final int originalHeight = decodeBoundsOptions.outHeight;
// inSampleSize prefers multiples of 2, but we prefer to prioritize memory savings
decodeBitmapOptions.inSampleSize= Math.max(1,Math.min(originalWidth / minimumDesiredBitmapWidth, originalHeight / minimumDesiredBitmapHeight));
}
return BitmapFactory.decodeStream(is,null,decodeBitmapOptions);
} catch( IOException e ) {
throw new RuntimeException(e); // this shouldn't happen
} finally {
try {
is.close();
} catch( IOException ignored ) {}
}
}

Here is my version, based on #emmby solution (thanks man!)
I've included a second phase where you take the reduced bitmap and scale it again to match exactly your desired dimensions.
My version takes a file path rather than a stream.
protected Bitmap createScaledBitmap(String filePath, int desiredBitmapWith, int desiredBitmapHeight) throws IOException, FileNotFoundException {
BufferedInputStream imageFileStream = new BufferedInputStream(new FileInputStream(filePath));
try {
// Phase 1: Get a reduced size image. In this part we will do a rough scale down
int sampleSize = 1;
if (desiredBitmapWith > 0 && desiredBitmapHeight > 0) {
final BitmapFactory.Options decodeBoundsOptions = new BitmapFactory.Options();
decodeBoundsOptions.inJustDecodeBounds = true;
imageFileStream.mark(64 * 1024);
BitmapFactory.decodeStream(imageFileStream, null, decodeBoundsOptions);
imageFileStream.reset();
final int originalWidth = decodeBoundsOptions.outWidth;
final int originalHeight = decodeBoundsOptions.outHeight;
// inSampleSize prefers multiples of 2, but we prefer to prioritize memory savings
sampleSize = Math.max(1, Math.max(originalWidth / desiredBitmapWith, originalHeight / desiredBitmapHeight));
}
BitmapFactory.Options decodeBitmapOptions = new BitmapFactory.Options();
decodeBitmapOptions.inSampleSize = sampleSize;
decodeBitmapOptions.inPreferredConfig = Bitmap.Config.RGB_565; // Uses 2-bytes instead of default 4 per pixel
// Get the roughly scaled-down image
Bitmap bmp = BitmapFactory.decodeStream(imageFileStream, null, decodeBitmapOptions);
// Phase 2: Get an exact-size image - no dimension will exceed the desired value
float ratio = Math.min((float)desiredBitmapWith/ (float)bmp.getWidth(), (float)desiredBitmapHeight/ (float)bmp.getHeight());
int w =(int) ((float)bmp.getWidth() * ratio);
int h =(int) ((float)bmp.getHeight() * ratio);
return Bitmap.createScaledBitmap(bmp, w,h, true);
} catch (IOException e) {
throw e;
} finally {
try {
imageFileStream.close();
} catch (IOException ignored) {
}
}
}