（一）AVFoundation-人脸检测，获取人脸图片并上传出

用到是AVFoundation 这个库

人脸检测使用思路：用Session管理，输入Input  ，呈现捕获对象Output，最后在layer上显示

---

使用到的如下类：

1.AVCaptureSession    继承与NSObject，是AVFoundation的核心类 ，用于管理捕获对象AVCaptureInput的视频和音频的输入，协调捕获的输出AVCaptureOutput

2.AVCaptureDeviceInput  用于从AVCaptureDevice对象捕获数据。

3.AVCaptureVideoDataOutput   继承于AVCaptureOutput 用于将捕获输出数据(如文件和视频预览)连接到捕获会话AVCaptureSession的实例 

4.AVCaptureMetadataOutput   继承于AVCaptureOutput （主要用于人脸检测）

---

定义对象：

@property (nonatomic,strong) AVCaptureSession *session;

@property (nonatomic,strong) AVCaptureDeviceInput*input；

@property (nonatomic, strong) AVCaptureMetadataOutput *MetadataOutput;

@property (nonatomic,strong) AVCaptureVideoDataOutput *videoOutput;

@property (nonatomic,strong) AVCaptureVideoPreviewLayer *previewLayer;

---

初始化各个对象：

-(void)deviceInit{

    //1.获取输入设备（摄像头）可以切换前后摄像头

    NSArray *devices = [AVCaptureDeviceDiscoverySession discoverySessionWithDeviceTypes:@[AVCaptureDeviceTypeBuiltInWideAngleCamera] mediaType:AVMediaTypeVideo position:AVCaptureDevicePositionBack].devices;

    AVCaptureDevice*deviceF = devices[0];

    //2.根据输入设备创建输入对象

    self.input= [[AVCaptureDeviceInputalloc]initWithDevice:deviceFerror:nil];

    // 设置代理监听输出对象输出的数据

    self.MetadataOutput = [[AVCaptureMetadataOutput alloc] init];

    // 设置代理监听输出对象输出的数据

    self.videoOutput = [[AVCaptureVideoDataOutput alloc] init];

    //对实时视频帧进行相关的渲染操作,指定代理

    [_videoOutput setSampleBufferDelegate:self queue:dispatch_get_main_queue()];

    self.session = [[AVCaptureSession alloc] init];

    //5.设置输出质量(高像素输出)

    if ([self.session canSetSessionPreset:AVCaptureSessionPreset640x480]) {

        [self.session setSessionPreset:AVCaptureSessionPreset640x480];

    }

    //6.添加输入和输出到会话

    [self.session beginConfiguration];

    if([self.sessioncanAddInput:_input]) {

        [self.sessionaddInput:_input];

    }

    if ([self.session canAddOutput:_MetadataOutput]) {

        [self.session addOutput:_MetadataOutput];

    }

    if ([self.session canAddOutput:_videoOutput]) {

           [self.session addOutput:_videoOutput];

       }

    [self.MetadataOutput setMetadataObjectTypes:@[AVMetadataObjectTypeFace]];

    [self.MetadataOutput setMetadataObjectsDelegate:self  queue:dispatch_queue_create("face", NULL)];

    self.MetadataOutput.rectOfInterest = self.view.bounds;

    [self.session commitConfiguration];

    AVCaptureSession *session = (AVCaptureSession *)self.session;

   //7.创建预览图层                                                                                                                         _previewLayer = [[AVCaptureVideoPreviewLayer alloc]   initWithSession:self.session];    _previewLayer.videoGravity = AVLayerVideoGravityResizeAspect;                                            _previewLayer.frame = self.view.bounds;                                                                                  [self.view.layer insertSublayer:_previewLayer atIndex:0];

    //8. 开始扫描

    [self.session startRunning];

}

主要用到这些代理

---

代理方法如下：

捕获到视频流，刷新率和手机的刷新率一样

AVCaptureVideoDataOutputSampleBufferDelegate

- (void)captureOutput:(AVCaptureOutput*)output didOutputSampleBuffer:(CMSampleBufferRef)sampleBuffer fromConnection:(AVCaptureConnection*)connection；

检测到人脸才会回调，只会返回一些人脸信息

AVCaptureMetadataOutputObjectsDelegate

- (void)captureOutput:(AVCaptureOutput*)outputdidOutputMetadataObjects:(NSArray<__kindofAVMetadataObject*> *)metadataObjectsfromConnection:(AVCaptureConnection*)connection;

识别到人脸后出现

  （ "",

    ""）