Taking photo snapshots requires that you first create a Player for
the capture action. Now you can realize the Player and get a VideoControl from
your Player's instance. In your application, you need to specify a locator
that is supported by your target device or platform. Supported Player locators
were discussed in section Recording
Sound and Video.
VideoControl has a method, getSnapshot,
which allows you to get a snapshot of the video it is displaying. This is
particularly useful when it is displaying live video from the device's camera,
since it allows you to take a photo. The following is a code sample of how
to take a photo using default settings:
byte[] pngImageData = videoControl.getSnapshot(null);
You can also specify other formats if the device supports them, and give a width and height:
videoControl.getSnapshot("encoding=bmp"); //
BMP, default size
videoControl.getSnapshot("width=80&height=60");
// default encoding, size 80x60
videoControl.getSnapshot("encoding=bmp&width=80&height=60");
// BMP, 80x60
If values are specified for width and height, both of them must be specified,
not just the width or height. In this case, the image is scaled to the requested
width and height. If the requested aspect ratio (that is, 4:3) does not match
that of the default size (in most devices 160 x 120), the resulting image
will be distorted, or in Series 40 devices, if the width and height do not
match a supported size, the closest match is used instead. The maximum size
that can be captured depends on the free heap memory available. The system
property video.snapshot.encodings contains a list of
the valid video snapshot encodings for your device. Supported image types
can be, for example, encoding=png, encoding=bmp, encoding=jpeg,
and encoding=gif. JPEG format is the most compact format.
The use of compact format reduces memory usage, and is therefore recommended
in many cases.
For a complete example of a MIDlet that takes photos using this method, see document Camera MIDlet: A Mobile Media Example.