CameraController

const val COORDINATE_SYSTEM_VIEW_REFERENCED = 1: Int

ImageAnalysis.Analyzer option for returning PreviewView coordinates.

When the ImageAnalysis.Analyzer is configured with this option, it will receive a Matrix that will receive a value that represents the transformation from camera sensor to the PreviewView, which can be used for highlighting detected result in PreviewView. For example, laying over a bounding box on top of the detected face.

Note this option only works if the ImageAnalysis.Analyzer is set via setImageAnalysisAnalyzer. It will not be effective when used with camera-core directly.

const val IMAGE_ANALYSIS = 2: Int

Bitmask option to enable ImageAnalysis. In setEnabledUseCases, if (enabledUseCases & IMAGE_ANALYSIS) != 0, then controller will enable image analysis features.

const val IMAGE_CAPTURE = 1: Int

Bitmask option to enable ImageCapture. In setEnabledUseCases, if (enabledUseCases & IMAGE_CAPTURE) != 0, then controller will enable image capture features.

const val TAP_TO_FOCUS_FAILED = 4: Int

The previous tap-to-focus action was failed to complete. This is usually due to device limitations.

const val TAP_TO_FOCUS_FOCUSED = 2: Int

The previous tap-to-focus action was completed successfully and the camera is focused.

const val TAP_TO_FOCUS_NOT_FOCUSED = 3: Int

The previous tap-to-focus action was completed successfully but the camera is still unfocused, similar to the CONTROL_AF_STATE_NOT_FOCUSED_LOCKED state. The end user might be able to get a better result by trying again with different camera distances and/or lighting.

const val TAP_TO_FOCUS_NOT_STARTED = 0: Int

No tap-to-focus action has been started by the end user.

const val TAP_TO_FOCUS_STARTED = 1: Int

A tap-to-focus action has started but not completed. The app also gets notified with this state if a new action happens before the previous one could finish.

const val VIDEO_CAPTURE = 4: Int

Bitmask option to enable video capture use case. In setEnabledUseCases, if (enabledUseCases & VIDEO_CAPTURE) != 0, then controller will enable video capture features.

@MainThread
fun clearEffects(): Unit

Removes all effects.

Once called, CameraX will remove all the effects and rebind the UseCase.

@MainThread
fun clearImageAnalysisAnalyzer(): Unit

Removes a previously set analyzer.

This will stop data from streaming to the ImageAnalysis.

If the current getDefaultTargetResolution returns non-null value, calling this method will reconfigure the camera which might cause additional latency. To avoid this, call this method when the lifecycle is not active.

@MainThread
fun enableTorch(torchEnabled: Boolean): ListenableFuture<Void!>

Enable the torch or disable the torch.

If the value is set before the camera is ready, CameraController waits for the camera to be ready and then enables the torch.

@MainThread
fun getCameraControl(): CameraControl?

Gets the CameraControl of the currently attached camera.

For controls available directly through CameraController as well as CameraControl, it's recommended to use the ones with CameraController, e.g. setLinearZoom v.s. setLinearZoom. CameraControl is a lower-layer API and may require more steps to achieve the same effect, and will not maintain control values when switching between cameras.

@MainThread
fun getCameraInfo(): CameraInfo?

Gets the CameraInfo of the currently attached camera.

For info available directly through CameraController as well as CameraInfo, it's recommended to use the ones with CameraController, e.g. getTorchState v.s. getTorchState. CameraInfo is a lower-layer API and may require more steps to achieve the same effect, and will not maintain values when switching between cameras.

@MainThread
fun getCameraSelector(): CameraSelector

Gets the CameraSelector.

The default value isDEFAULT_BACK_CAMERA.

@MainThread
fun getImageAnalysisBackgroundExecutor(): Executor?

Gets the default executor for ImageAnalysis background tasks.

@MainThread
fun getImageAnalysisBackpressureStrategy(): Int

Returns the mode with which images are acquired.

If not set, it defaults to STRATEGY_KEEP_ONLY_LATEST.

@MainThread
fun getImageAnalysisImageQueueDepth(): Int

Gets the image queue depth of ImageAnalysis.

@MainThread
fun getImageAnalysisOutputImageFormat(): Int

Gets the output image format for ImageAnalysis.

The returned image format can be either OUTPUT_IMAGE_FORMAT_YUV_420_888 or OUTPUT_IMAGE_FORMAT_RGBA_8888.

@MainThread
fun getImageAnalysisResolutionSelector(): ResolutionSelector?

Returns the ResolutionSelector for ImageAnalysis.

This method returns the value set by setImageAnalysisResolutionSelector. It returns null if the value has not been set.

@MainThread
fun getImageAnalysisTargetSize(): CameraController.OutputSize?

Returns the intended output size for ImageAnalysis set by setImageAnalysisTargetSize, or null if not set.

@MainThread
fun getImageCaptureFlashMode(): Int

Gets the flash mode for ImageCapture.

@MainThread
fun getImageCaptureIoExecutor(): Executor?

Gets the default executor for ImageCapture IO tasks.

@MainThread
fun getImageCaptureMode(): Int

Returns the image capture mode.

@MainThread
fun getImageCaptureResolutionSelector(): ResolutionSelector?

Returns the ResolutionSelector for ImageCapture.

This method returns the value set by setImageCaptureResolutionSelector (ResolutionSelector)}. It returns null if the value has not been set.

@MainThread
fun getImageCaptureTargetSize(): CameraController.OutputSize?

Returns the intended output size for ImageCapture set by setImageCaptureTargetSize, or null if not set.

fun getInitializationFuture(): ListenableFuture<Void!>

Gets a ListenableFuture that completes when camera initialization completes.

This future may fail with an InitializationException and associated cause that can be retrieved by getCause. The cause will be a CameraUnavailableException if it fails to access any camera during initialization.

In the rare case that the future fails with CameraUnavailableException, the camera will become unusable. This could happen for various reasons, for example hardware failure or the camera being held by another process. If the failure is temporary, killing and restarting the app might fix the issue.

The initialization also try to bind use cases before completing the ListenableFuture. The ListenableFuture will complete successfully regardless of whether the use cases are ready to be bound, e.g. it will complete successfully even if the controller is not set on a PreviewView. However the ListenableFuture will fail if the enabled use cases are not supported by the current camera.

@MainThread
fun getPreviewResolutionSelector(): ResolutionSelector?

Returns the ResolutionSelector for Preview.

This method returns the value set by setPreviewResolutionSelector. It returns null if the value has not been set.

@MainThread
fun getPreviewTargetSize(): CameraController.OutputSize?

Returns the intended output size for Preview set by setPreviewTargetSize, or null if not set.

@MainThread
fun getTapToFocusState(): LiveData<Int!>

Returns a LiveData with the latest tap-to-focus state.

When tap-to-focus feature is enabled, the LiveData will receive updates of focusing states. This happens when the end user taps on PreviewView, and then again when focusing is finished either successfully or unsuccessfully. The following table displays the states the LiveData can be in, and the possible transitions between them.

@MainThread
fun getTorchState(): LiveData<Int!>

Returns a LiveData of current TorchState.

The torch can be turned on and off via enableTorch which will trigger the change event to the returned LiveData.

@MainThread
fun getVideoCaptureDynamicRange(): DynamicRange

Gets the DynamicRange for video capture.

@MainThread
fun getVideoCaptureMirrorMode(): Int

Gets the mirror mode for video capture.

@MainThread
fun getVideoCaptureQualitySelector(): QualitySelector

Returns the QualitySelector for VIDEO_CAPTURE.

@MainThread
fun getVideoCaptureTargetFrameRate(): Range<Int!>

Gets the target frame rate in frames per second for video capture.

@MainThread
fun getZoomState(): LiveData<ZoomState!>

Returns a LiveData of ZoomState.

The LiveData will be updated whenever the set zoom state has been changed. This can occur when the application updates the zoom via setZoomRatio or setLinearZoom. The zoom state can also change anytime a camera starts up, for example when setCameraSelector is called.

@MainThread
fun hasCamera(cameraSelector: CameraSelector): Boolean

Checks if the given CameraSelector can be resolved to a camera.

Use this method to check if the device has the given camera.

Only call this method after camera is initialized. e.g. after the ListenableFuture from getInitializationFuture is finished. Calling it prematurely throws IllegalStateException. Example:

controller.getInitializationFuture().addListener(() -> {
    if (controller.hasCamera(cameraSelector)) {
        controller.setCameraSelector(cameraSelector);
    } else {
        // Update UI if the camera is not available.
    }
    // Attach PreviewView after we know the camera is available.
    previewView.setController(controller);
}, ContextCompat.getMainExecutor(requireContext()));

@MainThread
fun isImageAnalysisEnabled(): Boolean

Checks if ImageAnalysis is enabled.

@MainThread
fun isImageCaptureEnabled(): Boolean

Checks if ImageCapture is enabled.

ImageCapture is enabled by default. It has to be enabled before takePicture can be called.

@MainThread
fun isPinchToZoomEnabled(): Boolean

Returns whether pinch-to-zoom is enabled.

By default pinch-to-zoom is enabled.

@MainThread
fun isRecording(): Boolean

Returns whether there is an in-progress video recording.

@MainThread
fun isTapToFocusEnabled(): Boolean

Returns whether tap-to-focus is enabled.

By default tap-to-focus is enabled.

@MainThread
fun isVideoCaptureEnabled(): Boolean

Checks if video capture is enabled.

Video capture is disabled by default. It has to be enabled before startRecording can be called.

@MainThread
fun setCameraSelector(cameraSelector: CameraSelector): Unit

Sets the CameraSelector.

Calling this method with a CameraSelector that resolves to a different camera will change the camera being used by the controller. If camera initialization is complete, the controller will immediately rebind use cases with the new CameraSelector; otherwise, the new CameraSelector will be used when the camera becomes ready.

The default value is DEFAULT_BACK_CAMERA.

@MainThread
fun setEffects(effects: (Mutable)Set<CameraEffect!>): Unit

Sets CameraEffect.

Call this method to set a list of active effects. There is maximum one effect per UseCase. Adding effects with duplicate or invalid targets throws IllegalArgumentException. Once called, CameraX will rebind the UseCase with the effects applied. Effects not in the list are automatically removed.

The method throws IllegalArgumentException if the effects combination is not supported by CameraX. Please see the Javadoc of addEffect to see the supported effects combinations.

@MainThread
fun setEnabledUseCases(enabledUseCases: Int): Unit

Enables or disables use cases.

Use cases need to be enabled before they can be used. By default, IMAGE_CAPTURE and IMAGE_ANALYSIS are enabled, and VIDEO_CAPTURE is disabled. This is necessary because VIDEO_CAPTURE may affect the available resolutions for other use cases, especially on lower end devices.

To make sure getting the maximum resolution, only enable VIDEO_CAPTURE when shooting video. For example:

// By default, image capture is enabled. Taking picture works.
controller.takePicture(...);

// Switch to video capture to shoot video.
controller.setEnabledUseCases(VIDEO_CAPTURE);
controller.startRecording(...);

// Switch back to image capture and image analysis before taking another picture.
controller.setEnabledUseCases(IMAGE_CAPTURE|IMAGE_ANALYSIS);
controller.takePicture(...);

@MainThread
fun setImageAnalysisAnalyzer(
    executor: Executor,
    analyzer: ImageAnalysis.Analyzer
): Unit

Sets an analyzer to receive and analyze images.

Applications can process or copy the image by implementing the ImageAnalysis.Analyzer. The image needs to be closed by calling close when the analyzing is done.

Setting an analyzer function replaces any previous analyzer. Only one analyzer can be set at any time.

If the getTargetCoordinateSystem returns COORDINATE_SYSTEM_VIEW_REFERENCED, the analyzer will receive a transformation via updateTransform that converts coordinates from the ImageAnalysis's coordinate system to the PreviewView's coordinate system.

If the getDefaultTargetResolution returns a non-null value, calling this method will reconfigure the camera which might cause additional latency. To avoid this, set the value before controller is bound to the lifecycle.

@MainThread
fun setImageAnalysisBackgroundExecutor(executor: Executor?): Unit

Sets the executor that will be used for ImageAnalysis background tasks.

If not set, the background executor will default to an automatically generated Executor.

Changing the value will reconfigure the camera, which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setImageAnalysisBackpressureStrategy(strategy: Int): Unit

Sets the backpressure strategy to apply to the image producer to deal with scenarios where images may be produced faster than they can be analyzed.

The available values are STRATEGY_BLOCK_PRODUCER and STRATEGY_KEEP_ONLY_LATEST. If not set, the backpressure strategy will default to STRATEGY_KEEP_ONLY_LATEST.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setImageAnalysisImageQueueDepth(depth: Int): Unit

Sets the image queue depth of ImageAnalysis.

This sets the number of images available in parallel to ImageAnalysis.Analyzer. The value is only used if the backpressure strategy is STRATEGY_BLOCK_PRODUCER.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setImageAnalysisOutputImageFormat(imageAnalysisOutputImageFormat: Int): Unit

Sets the output image format for ImageAnalysis.

The supported output image format are OUTPUT_IMAGE_FORMAT_YUV_420_888 and OUTPUT_IMAGE_FORMAT_RGBA_8888.

If not set, OUTPUT_IMAGE_FORMAT_YUV_420_888 will be used.

Requesting OUTPUT_IMAGE_FORMAT_RGBA_8888 causes extra overhead because format conversion takes time.

Changing the value will reconfigure the camera, which may cause additional latency. To avoid this, set the value before controller is bound to lifecycle. If the value is changed when the camera is active, check the getFormat value to determine when the new format takes effect.

@MainThread
fun setImageAnalysisResolutionSelector(
    resolutionSelector: ResolutionSelector?
): Unit

Sets the ResolutionSelector for ImageAnalysis.

CameraX uses this value as a hint to select the resolution for images. The actual output may differ from the requested value due to device constraints. When set to null, CameraX will use the default config of ImageAnalysis. ImageAnalysis has a default ResolutionStrategy with bound size as 640x480 and fallback rule of FALLBACK_RULE_CLOSEST_HIGHER_THEN_LOWER.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setImageAnalysisTargetSize(targetSize: CameraController.OutputSize?): Unit

Sets the intended output size for ImageAnalysis.

The value is used as a hint when determining the resolution and aspect ratio of the output buffer. The actual output may differ from the requested value due to device constraints.

When set to null, the output will be based on the default config of ImageAnalysis.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setImageCaptureFlashMode(flashMode: Int): Unit

Sets the flash mode for ImageCapture.

If not set, the flash mode will default to FLASH_MODE_OFF.

If FLASH_MODE_SCREEN is set, a valid android.view.Window instance must be set to a PreviewView or ScreenFlashView which this controller is set to. Trying to use FLASH_MODE_SCREEN with a non-front camera or without setting a non-null window will be no-op. While switching the camera, it is the application's responsibility to change flash mode to the desired one if it leads to a no-op case (e.g. switching to rear camera while FLASH_MODE_SCREEN is still set). Otherwise, FLASH_MODE_OFF will be set.

@MainThread
fun setImageCaptureIoExecutor(executor: Executor?): Unit

Sets the default executor that will be used for ImageCapture IO tasks.

This executor will be used for any IO tasks specifically for ImageCapture, such as takePicture. If no executor is set, then a default Executor specifically for IO will be used instead.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setImageCaptureMode(captureMode: Int): Unit

Sets the image capture mode.

Valid capture modes are CAPTURE_MODE_MINIMIZE_LATENCY, which prioritizes latency over image quality, or CAPTURE_MODE_MAXIMIZE_QUALITY, which prioritizes image quality over latency.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setImageCaptureResolutionSelector(
    resolutionSelector: ResolutionSelector?
): Unit

Sets the ResolutionSelector for ImageCapture.

CameraX uses this value as a hint to select the resolution for captured images. The actual output may differ from the requested value due to device constraints. When set to null, CameraX will use the default config of ImageCapture. The default resolution strategy for ImageCapture is HIGHEST_AVAILABLE_STRATEGY, which will select the largest available resolution to use.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setImageCaptureTargetSize(targetSize: CameraController.OutputSize?): Unit

Sets the intended image size for ImageCapture.

The value is used as a hint when determining the resolution and aspect ratio of the captured image. The actual output may differ from the requested value due to device constraints.

When set to null, the output will be based on the default config of ImageCapture.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setLinearZoom(linearZoom: @FloatRange(from = 0.0, to = 1.0) Float): ListenableFuture<Void!>

Sets current zoom by a linear zoom value ranging from 0f to 1.0f.

LinearZoom 0f represents the minimum zoom while linearZoom 1.0f represents the maximum zoom. The advantage of linearZoom is that it ensures the field of view (FOV) varies linearly with the linearZoom value, for use with slider UI elements (while setZoomRatio works well for pinch-zoom gestures).

If the value is set before the camera is ready, CameraController waits for the camera to be ready and then sets the linear zoom.

@MainThread
fun setPinchToZoomEnabled(enabled: Boolean): Unit

Enables/disables pinch-to-zoom.

Once enabled, end user can pinch on the PreviewView to zoom in/out if the bound camera supports zooming.

@MainThread
fun setPreviewResolutionSelector(resolutionSelector: ResolutionSelector?): Unit

Sets the ResolutionSelector for Preview.

CameraX uses this value as a hint to select the resolution for preview. The actual output may differ from the requested value due to device constraints. When set to null, CameraX will use the default config of Preview. By default, the selected resolution will be limited by the PREVIEW size which is defined as the best size match to the device's screen resolution, or to 1080p (1920x1080), whichever is smaller.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setPreviewTargetSize(targetSize: CameraController.OutputSize?): Unit

Sets the intended output size for Preview.

The value is used as a hint when determining the resolution and aspect ratio of the preview. The actual output may differ from the requested value due to device constraints.

When set to null, the output will be based on the default config of Preview.

Changing the value will reconfigure the camera which will cause additional latency. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setTapToFocusEnabled(enabled: Boolean): Unit

Enables/disables tap-to-focus.

Once enabled, end user can tap on the PreviewView to set focus point.

@MainThread
fun setVideoCaptureDynamicRange(dynamicRange: DynamicRange): Unit

Sets the DynamicRange for video capture.

The dynamic range specifies how the range of colors, highlights and shadows that are captured by the video producer are displayed on a display. Some dynamic ranges will allow the video to make full use of the extended range of brightness of a display when the video is played back.

The supported dynamic ranges for video capture can be queried through the androidx.camera.video.VideoCapabilities returned by getVideoCapabilities via getSupportedDynamicRanges.

It is possible to choose a high dynamic range (HDR) with unspecified encoding by providing HDR_UNSPECIFIED_10_BIT.

If the dynamic range is not provided, the default value is SDR.

@MainThread
fun setVideoCaptureMirrorMode(mirrorMode: Int): Unit

Sets the mirror mode for video capture.

Valid values include: MIRROR_MODE_OFF, MIRROR_MODE_ON and MIRROR_MODE_ON_FRONT_ONLY. If not set, it defaults to MIRROR_MODE_OFF.

@MainThread
fun setVideoCaptureQualitySelector(qualitySelector: QualitySelector): Unit

Sets the QualitySelector for VIDEO_CAPTURE.

The provided quality selector is used to select the resolution of the recording depending on the resolutions supported by the camera and codec capabilities.

If no quality selector is provided, the default is DEFAULT_QUALITY_SELECTOR.

Changing the value will reconfigure the camera which will cause video capture to stop. To avoid this, set the value before controller is bound to lifecycle.

@MainThread
fun setVideoCaptureTargetFrameRate(targetFrameRate: Range<Int!>): Unit

Sets the target frame rate range in frames per second for video capture.

This target will be used as a part of the heuristics for the algorithm that determines the final frame rate range and resolution of all concurrently bound use cases.

It is not guaranteed that this target frame rate will be the final range, as other use cases as well as frame rate restrictions of the device may affect the outcome of the algorithm that chooses the actual frame rate.

By default, the value is FRAME_RATE_RANGE_UNSPECIFIED. For supported frame rates, see getSupportedFrameRateRanges.

@MainThread
fun setZoomRatio(zoomRatio: Float): ListenableFuture<Void!>

Sets current zoom by ratio.

Valid zoom values range from getMinZoomRatio to getMaxZoomRatio.

If the value is set before the camera is ready, CameraController waits for the camera to be ready and then sets the zoom ratio.

@RequiresApi(value = 26)
@MainThread
fun startRecording(
    outputOptions: FileDescriptorOutputOptions,
    audioConfig: AudioConfig,
    executor: Executor,
    listener: Consumer<VideoRecordEvent!>
): Recording

Takes a video to a given file descriptor.

Currently, file descriptors as output destinations are not supported on pre-Android O (API 26) devices.

Only a single recording can be active at a time, so if isRecording is true, this will throw an IllegalStateException.

Upon successfully starting the recording, a VideoRecordEvent.Start event will be the first event sent to the provided event listener.

If errors occur while starting the recording, a VideoRecordEvent.Finalize event will be the first event sent to the provided listener, and information about the error can be found in that event's getError method.

Recording with audio requires the RECORD_AUDIO permission; without it, starting a recording will fail with a SecurityException.

@MainThread
fun startRecording(
    outputOptions: FileOutputOptions,
    audioConfig: AudioConfig,
    executor: Executor,
    listener: Consumer<VideoRecordEvent!>
): Recording

Takes a video to a given file.

Only a single recording can be active at a time, so if isRecording is true, this will throw an IllegalStateException.

Upon successfully starting the recording, a VideoRecordEvent.Start event will be the first event sent to the provided event listener.

If errors occur while starting the recording, a VideoRecordEvent.Finalize event will be the first event sent to the provided listener, and information about the error can be found in that event's getError method.

Recording with audio requires the RECORD_AUDIO permission; without it, starting a recording will fail with a SecurityException.

@MainThread
fun startRecording(
    outputOptions: MediaStoreOutputOptions,
    audioConfig: AudioConfig,
    executor: Executor,
    listener: Consumer<VideoRecordEvent!>
): Recording

Takes a video to MediaStore.

Only a single recording can be active at a time, so if isRecording is true, this will throw an IllegalStateException.

Upon successfully starting the recording, a VideoRecordEvent.Start event will be the first event sent to the provided event listener.

If errors occur while starting the recording, a VideoRecordEvent.Finalize event will be the first event sent to the provided listener, and information about the error can be found in that event's getError method.

Recording with audio requires the RECORD_AUDIO permission; without it, starting a recording will fail with a SecurityException.

@MainThread
fun takePicture(
    executor: Executor,
    callback: ImageCapture.OnImageCapturedCallback
): Unit

Captures a new still image for in memory access.

The listener is responsible for calling close on the returned image.

@MainThread
fun takePicture(
    outputFileOptions: ImageCapture.OutputFileOptions,
    executor: Executor,
    imageSavedCallback: ImageCapture.OnImageSavedCallback
): Unit

Captures a new still image and saves to a file along with application specified metadata.

The callback will be called only once for every invocation of this method.

By default, the saved image is mirrored to match the output of the preview if front camera is used. To override this behavior, the app needs to explicitly set the flag to false using setReversedHorizontal and setMetadata.

The saved image is cropped to match the aspect ratio of the PreviewView. To take a picture with the maximum available resolution, make sure that the PreviewView's aspect ratio matches the max JPEG resolution supported by the camera.