iphone - tipos - AVFoundation toque para enfocar el rectángulo de retroalimentación

retroalimentacion formativa

Question

Estoy desarrollando una aplicación para iPhone en la que utilizo directamente AVFoundation para capturar videos a través de la cámara.

He implementado una función para habilitar la función de tap to focus para un usuario.

- (void) focus:(CGPoint) aPoint;
{
#if HAS_AVFF
    Class captureDeviceClass = NSClassFromString(@"AVCaptureDevice");
    if (captureDeviceClass != nil) {        
        AVCaptureDevice *device = [captureDeviceClass defaultDeviceWithMediaType:AVMediaTypeVideo];
        if([device isFocusPointOfInterestSupported] &&
           [device isFocusModeSupported:AVCaptureFocusModeAutoFocus]) {
            CGRect screenRect = [[UIScreen mainScreen] bounds];
            double screenWidth = screenRect.size.width;
            double screenHeight = screenRect.size.height;
            double focus_x = aPoint.x/screenWidth;
            double focus_y = aPoint.y/screenHeight;
            if([device lockForConfiguration:nil]) {
                [device setFocusPointOfInterest:CGPointMake(focus_x,focus_y)];
                [device setFocusMode:AVCaptureFocusModeAutoFocus];
                if ([device isExposureModeSupported:AVCaptureExposureModeAutoExpose]){
                    [device setExposureMode:AVCaptureExposureModeAutoExpose];
                }
                [device unlockForConfiguration];
            }
        }
    }
#endif
}

Hasta ahora todo bien, pero me falta el rectángulo de comentarios como en la aplicación de fotos. ¿Hay alguna manera de decirle a AVFoundation Framework que muestre este rectángulo de comentarios o tengo que implementar esta función yo mismo?

Answer 1

Agregando a la brillante respuesta de Anil: en lugar de hacer los cálculos usted mismo, debería echar un vistazo a la captura de dispositivo de AVCaptureVideoPreviewLayer ''s captureDevicePointOfInterestForPoint:. Le dará un punto de enfoque mucho más consistente (disponible desde iOS 6 y versiones posteriores).

- (void) focus:(CGPoint) aPoint;
{
    Class captureDeviceClass = NSClassFromString(@"AVCaptureDevice");
    if (captureDeviceClass != nil) {
        AVCaptureDevice *device = [captureDeviceClass defaultDeviceWithMediaType:AVMediaTypeVideo];
        if([device isFocusPointOfInterestSupported] &&
           [device isFocusModeSupported:AVCaptureFocusModeAutoFocus]) {

            CGPoint focusPoint = [self.captureVideoPreviewLayer captureDevicePointOfInterestForPoint:aPoint];
            if([device lockForConfiguration:nil]) {
                [device setFocusPointOfInterest:CGPointMake(focusPoint.x,focusPoint.y)];
                [device setFocusMode:AVCaptureFocusModeAutoFocus];
                if ([device isExposureModeSupported:AVCaptureExposureModeAutoExpose]){
                    [device setExposureMode:AVCaptureExposureModeAutoExpose];
                }
                [device unlockForConfiguration];
            }
        }
    }
}

La documentación está disponible aquí: https://developer.apple.com/library/ios/documentation/AVFoundation/Reference/AVCaptureVideoPreviewLayer_Class/index.html#//apple_ref/occ/instm/AVCaptureVideoPreviewLayer/captureDevicePointOfInterestForPoint .

Answer 2

Esto es lo que hice: esta es la clase que crea el cuadrado que se muestra cuando el usuario toca la superposición de la cámara.

CameraFocusSquare.h

#import <UIKit/UIKit.h>
@interface CameraFocusSquare : UIView
@end


CameraFocusSquare.m

#import "CameraFocusSquare.h"
#import <QuartzCore/QuartzCore.h>

const float squareLength = 80.0f;
@implementation FBKCameraFocusSquare

- (id)initWithFrame:(CGRect)frame
{
    self = [super initWithFrame:frame];
    if (self) {
        // Initialization code

        [self setBackgroundColor:[UIColor clearColor]];
        [self.layer setBorderWidth:2.0];
        [self.layer setCornerRadius:4.0];
        [self.layer setBorderColor:[UIColor whiteColor].CGColor];

        CABasicAnimation* selectionAnimation = [CABasicAnimation
                                                animationWithKeyPath:@"borderColor"];
        selectionAnimation.toValue = (id)[UIColor blueColor].CGColor;
        selectionAnimation.repeatCount = 8;
        [self.layer addAnimation:selectionAnimation
                          forKey:@"selectionAnimation"];

    }
    return self;
}
@end

Y en la vista donde recibe los toques, haga lo siguiente:

- (void) touchesBegan:(NSSet *)touches withEvent:(UIEvent *)event
{
    UITouch *touch = [[event allTouches] anyObject];
    CGPoint touchPoint = [touch locationInView:touch.view];
    [self focus:touchPoint];

    if (camFocus)
    {
        [camFocus removeFromSuperview];
    }
    if ([[touch view] isKindOfClass:[FBKVideoRecorderView class]])
    {
        camFocus = [[CameraFocusSquare alloc]initWithFrame:CGRectMake(touchPoint.x-40, touchPoint.y-40, 80, 80)];
        [camFocus setBackgroundColor:[UIColor clearColor]];
        [self addSubview:camFocus];
        [camFocus setNeedsDisplay];

        [UIView beginAnimations:nil context:NULL];
        [UIView setAnimationDuration:1.5];
        [camFocus setAlpha:0.0];
        [UIView commitAnimations];
    }
}

- (void) focus:(CGPoint) aPoint;
{
    Class captureDeviceClass = NSClassFromString(@"AVCaptureDevice");
    if (captureDeviceClass != nil) {
        AVCaptureDevice *device = [captureDeviceClass defaultDeviceWithMediaType:AVMediaTypeVideo];
        if([device isFocusPointOfInterestSupported] &&
           [device isFocusModeSupported:AVCaptureFocusModeAutoFocus]) {
            CGRect screenRect = [[UIScreen mainScreen] bounds];
            double screenWidth = screenRect.size.width;
            double screenHeight = screenRect.size.height;
            double focus_x = aPoint.x/screenWidth;
            double focus_y = aPoint.y/screenHeight;
            if([device lockForConfiguration:nil]) {
                [device setFocusPointOfInterest:CGPointMake(focus_x,focus_y)];
                [device setFocusMode:AVCaptureFocusModeAutoFocus];
                if ([device isExposureModeSupported:AVCaptureExposureModeAutoExpose]){
                    [device setExposureMode:AVCaptureExposureModeAutoExpose];
                }
                [device unlockForConfiguration];
            }
        }
    }
}

Answer 3

La respuesta de @ Anil es un gran comienzo, pero no funcionó para mí. Quería poder hacer que el usuario continúe seleccionando un punto de enfoque, en lugar de solo una vez (que es lo que hace su solución). Gracias a @Anil por señalarme en la dirección correcta.

Hay algunas diferencias con mi solución.

1. Quería poder reutilizar el cuadro de enfoque y la animación, en lugar de solo una vez.
2. Quería que la animación desapareciera después de que se completara (algo que no pude conseguir con la solución de @ Anil).
3. En lugar de usar initWithFrame: implementé mi propio initWithTouchPoint:
4. Tengo un método para animar específicamente la acción de enfoque.
5. También tengo un método para actualizar la ubicación del marco.
6. El tamaño del marco está dentro de CameraFocusSquare , lo que significa que es más fácil encontrar y actualizar el tamaño según sea necesario.

CameraFocusSquare.h

@import UIKit;

@interface CameraFocusSquare : UIView

- (instancetype)initWithTouchPoint:(CGPoint)touchPoint;
- (void)updatePoint:(CGPoint)touchPoint;
- (void)animateFocusingAction;

@end

CameraFocusSquare.m

#import "CameraFocusSquare.h"

@implementation CameraFocusSquare {
    CABasicAnimation *_selectionBlink;
}

/**
 This is the init method for the square. It sets the frame for the view and sets border parameters. It also creates the blink animation.
 */
- (instancetype)initWithTouchPoint:(CGPoint)touchPoint {
    self = [self init];
    if (self) {
        [self updatePoint:touchPoint];
        self.backgroundColor = [UIColor clearColor];
        self.layer.borderWidth = 2.0f;
        self.layer.borderColor = [UIColor orangeColor].CGColor;

        // create the blink animation
        _selectionBlink = [CABasicAnimation
                animationWithKeyPath:@"borderColor"];
        _selectionBlink.toValue = (id)[UIColor whiteColor].CGColor;
        _selectionBlink.repeatCount = 3;  // number of blinks
        _selectionBlink.duration = 0.4;  // this is duration per blink
        _selectionBlink.delegate = self;
    }
    return self;
}

/**
 Updates the location of the view based on the incoming touchPoint.
 */
- (void)updatePoint:(CGPoint)touchPoint {
    CGFloat squareWidth = 50;
    CGRect frame = CGRectMake(touchPoint.x - squareWidth/2, touchPoint.y - squareWidth/2, squareWidth, squareWidth);
    self.frame = frame;
}

/**
 This unhides the view and initiates the animation by adding it to the layer.
 */
- (void)animateFocusingAction {
    // make the view visible
    self.alpha = 1.0f;
    self.hidden = NO;
    // initiate the animation
    [self.layer addAnimation:_selectionBlink forKey:@"selectionAnimation"];
}

/**
 Hides the view after the animation stops. Since the animation is automatically removed, we don''t need to do anything else here.
 */
- (void)animationDidStop:(CAAnimation *)animation finished:(BOOL)flag {
    // hide the view
    self.alpha = 0.0f;
    self.hidden = YES;
}

@end

Inicié todo esto encima de una vista. Esto me permite una mayor flexibilidad y separa mi código de UI de mi código de controlador (piense en MVC).

PreviewView.h

@import UIKit;

@interface PreviewView : UIView

- (IBAction)tapToFocus:(UITapGestureRecognizer *)gestureRecognizer;

@end

Vista previa.m

#import "PreviewView.h"
#import "CameraFocusSquare.h"

@implementation PreviewView {
    CameraFocusSquare *_focusSquare;
}

- (IBAction)tapToFocus:(UITapGestureRecognizer *)gestureRecognizer {
    CGPoint touchPoint = [gestureRecognizer locationOfTouch:0 inView:self];
    if (!_focusSquare) {
        _focusSquare = [[CameraFocusSquare alloc] initWithTouchPoint:touchPoint];
        [self addSubview:_focusSquare];
        [_focusSquare setNeedsDisplay];
    }
    else {
        [_focusSquare updatePoint:touchPoint];
    }
    [_focusSquare animateFocusingAction];
}

@end

Finalmente, en mi subclase UIViewController , tengo mi UITapGestureRecognizer creado y adjunto a la vista. También implemento mi código de toque para enfocar aquí.

CameraViewController.m

- (void)viewDidLoad {
    // do other initialization stuff here

    // create the tap-to-focus gesture
    UITapGestureRecognizer *tapToFocusRecognizer = [[UITapGestureRecognizer alloc] initWithTarget:self action:@selector(tapToFocus:)];
    tapToFocusRecognizer.numberOfTapsRequired = 1;
    tapToFocusRecognizer.numberOfTouchesRequired = 1;
    [self.previewView addGestureRecognizer:tapToFocusRecognizer];
}

- (IBAction)tapToFocus:(UITapGestureRecognizer *)tapGestureRecognizer {
    if (!_captureDevice) {
        return;
    }
    if (![_captureDevice isFocusPointOfInterestSupported]) {
        return;
    }
    if (![_captureDevice isFocusModeSupported:AVCaptureFocusModeAutoFocus]) {
        return;
    }
    [self.previewView tapToFocus:tapGestureRecognizer];
    NSError *error;
    [_captureDevice lockForConfiguration:&error];
    if (error) {
        NSLog(@"Error trying to lock configuration of camera. %@", [error localizedDescription]);
        return;
    }
    CGPoint touchPoint = [tapGestureRecognizer locationOfTouch:0 inView:self.cameraView];
    // range of touch point is from (0,0) to (1,1)
    CGFloat touchX = touchPoint.x / self.previewView.frame.size.width;
    CGFloat touchY = touchPoint.y / self.previewView.frame.size.height;

    _captureDevice.focusMode = AVCaptureFocusModeAutoFocus;
    if ([_captureDevice isExposureModeSupported:AVCaptureExposureModeAutoExpose]) {
        _captureDevice.exposureMode = AVCaptureExposureModeAutoExpose;
    }
    _captureDevice.focusPointOfInterest = CGPointMake(touchX, touchY);
    if ([_captureDevice isExposurePointOfInterestSupported]) {
        _captureDevice.exposurePointOfInterest = CGPointMake(touchX, touchY);
    }
    [_captureDevice unlockForConfiguration];
}

¡Espero que esto ayude a las personas para que puedan pasar a un código más importante!

Answer 4

Implementación rápida:

Vista de CameraFocusSquare:

class CameraFocusSquare: UIView,CAAnimationDelegate {

internal let kSelectionAnimation:String = "selectionAnimation"

fileprivate var _selectionBlink: CABasicAnimation?

convenience init(touchPoint: CGPoint) {
    self.init()
    self.updatePoint(touchPoint)
    self.backgroundColor = UIColor.clear
    self.layer.borderWidth = 2.0
    self.layer.borderColor = UIColor.orange.cgColor
    initBlink()
}

override init(frame: CGRect) {
    super.init(frame: frame)
}

fileprivate func initBlink() {
    // create the blink animation
    self._selectionBlink = CABasicAnimation(keyPath: "borderColor")
    self._selectionBlink!.toValue = (UIColor.white.cgColor as AnyObject)
    self._selectionBlink!.repeatCount = 3
    // number of blinks
    self._selectionBlink!.duration = 0.4
    // this is duration per blink
    self._selectionBlink!.delegate = self
}



required init?(coder aDecoder: NSCoder) {
    fatalError("init(coder:) has not been implemented")
}

/**
 Updates the location of the view based on the incoming touchPoint.
 */

func updatePoint(_ touchPoint: CGPoint) {
    let squareWidth: CGFloat = 100
    let frame: CGRect = CGRect(x: touchPoint.x - squareWidth / 2, y: touchPoint.y - squareWidth / 2, width: squareWidth, height: squareWidth)
    self.frame = frame
}
/**
 This unhides the view and initiates the animation by adding it to the layer.
 */

func animateFocusingAction() {

    if let blink = _selectionBlink {
        // make the view visible
        self.alpha = 1.0
        self.isHidden = false
        // initiate the animation
        self.layer.add(blink, forKey: kSelectionAnimation)
    }

}
/**
 Hides the view after the animation stops. Since the animation is automatically removed, we don''t need to do anything else here.
 */

public func animationDidStop(_ anim: CAAnimation, finished flag: Bool){
    if flag{
        // hide the view
        self.alpha = 0.0
        self.isHidden = true
    }
}

}

Acción gestual:

 open func tapToFocus(_ gesture : UILongPressGestureRecognizer) {

    if (gesture.state == UIGestureRecognizerState.began) {

        let touchPoint:CGPoint = gesture.location(in: self.previewView)

        if let fsquare = self.focusSquare {
            fsquare.updatePoint(touchPoint)
        }else{
            self.focusSquare = CameraFocusSquare(touchPoint: touchPoint)
            self.previewView.addSubview(self.focusSquare!)
            self.focusSquare?.setNeedsDisplay()
        }

        self.focusSquare?.animateFocusingAction()

        let convertedPoint:CGPoint = self.previewLayer!.captureDevicePointOfInterest(for: touchPoint)

        let currentDevice:AVCaptureDevice = self.videoDeviceInput!.device

        if currentDevice.isFocusPointOfInterestSupported && currentDevice.isFocusModeSupported(AVCaptureFocusMode.autoFocus){

            do {

                try currentDevice.lockForConfiguration()
                currentDevice.focusPointOfInterest = convertedPoint
                currentDevice.focusMode = AVCaptureFocusMode.autoFocus

                if currentDevice.isExposureModeSupported(AVCaptureExposureMode.continuousAutoExposure){
                    currentDevice.exposureMode = AVCaptureExposureMode.continuousAutoExposure
                }
                currentDevice.isSubjectAreaChangeMonitoringEnabled = true
                currentDevice.unlockForConfiguration()

            } catch {

            }
        }
    }
}