Merge \\"media: document output image crop and layout for MediaCodec\\" into nyc-dev am: 939484b8ab
am: 69484dcd8a
Change-Id: I12c8a8673ae7377bc9be5b0b6fc78a373eed97ee
This commit is contained in:
@@ -125,6 +125,77 @@ import java.util.Map;
|
||||
All video codecs support flexible YUV 4:2:0 buffers since {@link
|
||||
android.os.Build.VERSION_CODES#LOLLIPOP_MR1}.
|
||||
|
||||
<h4>Accessing Raw Video ByteBuffers on Older Devices</h4>
|
||||
<p>
|
||||
Prior to {@link android.os.Build.VERSION_CODES#LOLLIPOP} and {@link Image} support, you need to
|
||||
use the {@link MediaFormat#KEY_STRIDE} and {@link MediaFormat#KEY_SLICE_HEIGHT} output format
|
||||
values to understand the layout of the raw output buffers.
|
||||
<p class=note>
|
||||
Note that on some devices the slice-height is advertised as 0. This could mean either that the
|
||||
slice-height is the same as the frame height, or that the slice-height is the frame height
|
||||
aligned to some value (usually a power of 2). Unfortunately, there is no way to tell the actual
|
||||
slice height in this case. Furthermore, the vertical stride of the {@code U} plane in planar
|
||||
formats is also not specified or defined, though usually it is half of the slice height.
|
||||
<p>
|
||||
The {@link MediaFormat#KEY_WIDTH} and {@link MediaFormat#KEY_HEIGHT} keys specify the size of the
|
||||
video frames; however, for most encondings the video (picture) only occupies a portion of the
|
||||
video frame. This is represented by the 'crop rectangle'.
|
||||
<p>
|
||||
You need to use the following keys to get the crop rectangle of raw output images from the
|
||||
{@linkplain #getOutputFormat output format}. If these keys are not present, the video occupies the
|
||||
entire video frame.The crop rectangle is understood in the context of the output frame
|
||||
<em>before</em> applying any {@linkplain MediaFormat#KEY_ROTATION rotation}.
|
||||
<table style="width: 0%">
|
||||
<thead>
|
||||
<tr>
|
||||
<th>Format Key</th>
|
||||
<th>Type</th>
|
||||
<th>Description</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr>
|
||||
<td>{@code "crop-left"}</td>
|
||||
<td>Integer</td>
|
||||
<td>The left-coordinate (x) of the crop rectangle</td>
|
||||
</tr><tr>
|
||||
<td>{@code "crop-top"}</td>
|
||||
<td>Integer</td>
|
||||
<td>The top-coordinate (y) of the crop rectangle</td>
|
||||
</tr><tr>
|
||||
<td>{@code "crop-right"}</td>
|
||||
<td>Integer</td>
|
||||
<td>The right-coordinate (x) <strong>MINUS 1</strong> of the crop rectangle</td>
|
||||
</tr><tr>
|
||||
<td>{@code "crop-bottom"}</td>
|
||||
<td>Integer</td>
|
||||
<td>The bottom-coordinate (y) <strong>MINUS 1</strong> of the crop rectangle</td>
|
||||
</tr><tr>
|
||||
<td colspan=3>
|
||||
The right and bottom coordinates can be understood as the coordinates of the right-most
|
||||
valid column/bottom-most valid row of the cropped output image.
|
||||
</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<p>
|
||||
The size of the video frame (before rotation) can be calculated as such:
|
||||
<pre class=prettyprint>
|
||||
MediaFormat format = decoder.getOutputFormat(…);
|
||||
int width = format.getInteger(MediaFormat.KEY_WIDTH);
|
||||
if (format.containsKey("crop-left") && format.containsKey("crop-right")) {
|
||||
width = format.getInteger("crop-right") + 1 - format.getInteger("crop-left");
|
||||
}
|
||||
int height = format.getInteger(MediaFormat.KEY_HEIGHT);
|
||||
if (format.containsKey("crop-top") && format.containsKey("crop-bottom")) {
|
||||
height = format.getInteger("crop-bottom") + 1 - format.getInteger("crop-top");
|
||||
}
|
||||
</pre>
|
||||
<p class=note>
|
||||
Also note that the meaning of {@link BufferInfo#offset BufferInfo.offset} was not consistent across
|
||||
devices. On some devices the offset pointed to the top-left pixel of the crop rectangle, while on
|
||||
most devices it pointed to the top-left pixel of the entire frame.
|
||||
|
||||
<h3>States</h3>
|
||||
<p>
|
||||
During its life a codec conceptually exists in one of three states: Stopped, Executing or
|
||||
|
||||
@@ -295,17 +295,19 @@ public final class MediaFormat {
|
||||
* Stride (or row increment) is the difference between the index of a pixel
|
||||
* and that of the pixel directly underneath. For YUV 420 formats, the
|
||||
* stride corresponds to the Y plane; the stride of the U and V planes can
|
||||
* be calculated based on the color format.
|
||||
* be calculated based on the color format, though it is generally undefined
|
||||
* and depends on the device and release.
|
||||
* The associated value is an integer, representing number of bytes.
|
||||
*/
|
||||
public static final String KEY_STRIDE = "stride";
|
||||
|
||||
/**
|
||||
* A key describing the plane height of a multi-planar (YUV) video bytebuffer layout.
|
||||
* Slice height (or plane height) is the number of rows that must be skipped to get
|
||||
* from the top of the Y plane to the top of the U plane in the bytebuffer. In essence
|
||||
* Slice height (or plane height/vertical stride) is the number of rows that must be skipped
|
||||
* to get from the top of the Y plane to the top of the U plane in the bytebuffer. In essence
|
||||
* the offset of the U plane is sliceHeight * stride. The height of the U/V planes
|
||||
* can be calculated based on the color format.
|
||||
* can be calculated based on the color format, though it is generally undefined
|
||||
* and depends on the device and release.
|
||||
* The associated value is an integer, representing number of rows.
|
||||
*/
|
||||
public static final String KEY_SLICE_HEIGHT = "slice-height";
|
||||
|
||||
Reference in New Issue
Block a user