iOS 如何在使用 AVAssetWriter 捕获视频时正确处理方向

标签 ios swift video avassetwriter avassetwriterinput

我正在制作一个利用 AVFoundation 录制视频的示例应用程序。关键是我可以更好地控制视频的录制方式。在我的示例项目中,我进行了视频捕获,但在正确处理方向方面遇到困难。

我在网络上进行了大量搜索,发现其他人建议我不应该允许我的捕获 View 或捕获 session 根据方向旋转,而是设置一个转换以在播放期间旋转视频。我在 iOS 和 Mac 设备上运行良好,但想知道我是否会在其他平台(如 Windows 或 Android)上遇到问题。

此外,当我查看录制视频的元数据时,我发现宽度和高度未根据方向正确设置。这是有道理的,因为我只是在转换视频的呈现方式,而不是它的实际分辨率。

我的问题是如何正确支持纵向和横向方向并将其正确反射(reflect)在视频文件输出中?我需要这些视频在所有平台上正确播放,所以我认为分辨率非常重要。

以下是我迄今为止编写的完整源代码。感谢大家提供的任何建议。

import UIKit
import AVFoundation

class ViewController: UIViewController, AVCaptureVideoDataOutputSampleBufferDelegate, AVCaptureAudioDataOutputSampleBufferDelegate {

    //MARK: - Outlet

    @IBOutlet weak var previewView: UIView!
    @IBOutlet var playStopButton: UIButton!

    //MARK: - Private Variables

    let videoQueue = DispatchQueue(label: "VideoQueue", qos: .background, attributes: .concurrent, autoreleaseFrequency: .workItem, target: nil)
    private let captureSession = AVCaptureSession()

    var outputUrl: URL {
        get {

            if let url = _outputUrl {
                return url
            }


            _outputUrl = outputDirectory.appendingPathComponent("video.mp4")
            return _outputUrl!
        }
    }

    private var _outputUrl: URL?

    var outputDirectory: URL {
        get {

            if let url = _outputDirectory {
                return url
            }


            _outputDirectory = getDocumentsDirectory().appendingPathComponent("recording")
            return _outputDirectory!
        }
    }

    private var _outputDirectory: URL?

    private var assetWriter: AVAssetWriter?
    private var videoInput: AVAssetWriterInput?
    private var audioInput: AVAssetWriterInput?
    private var videoOutput: AVCaptureVideoDataOutput?
    private var audioOutput: AVCaptureAudioDataOutput?

    private var isRecording = false
    private var isWriting = false

    private var videoSize = CGSize(width: 640, height: 480)

    //MARK: - View Life-cycle

    override func viewDidLoad() {
        super.viewDidLoad()


        videoQueue.async {

            do {

                try self.configureCaptureSession()
                try self.configureAssetWriter()

                DispatchQueue.main.async {
                    self.configurePreview()
                }

            } catch {

                DispatchQueue.main.async {
                    self.showAlert("Unable to configure video output")
                }
            }
        }
    }

    override var preferredInterfaceOrientationForPresentation: UIInterfaceOrientation {
        return .portrait
    }

    override var supportedInterfaceOrientations: UIInterfaceOrientationMask {
        return .portrait
    }

    //MARK: - Capture Session

    private func configureCaptureSession() throws {

        do {

            // configure the session
            if captureSession.canSetSessionPreset(AVCaptureSessionPreset640x480) {
                captureSession.sessionPreset = AVCaptureSessionPreset640x480
            }

            // configure capture devices
            let camDevice = AVCaptureDevice.defaultDevice(withMediaType: AVMediaTypeVideo)
            let micDevice = AVCaptureDevice.defaultDevice(withMediaType: AVMediaTypeAudio)

            let camInput = try AVCaptureDeviceInput(device: camDevice)
            let micInput = try AVCaptureDeviceInput(device: micDevice)

            if captureSession.canAddInput(camInput) {
                captureSession.addInput(camInput)
            }

            if captureSession.canAddInput(micInput) {
                captureSession.addInput(micInput)
            }

            // configure audio/video output
            videoOutput = AVCaptureVideoDataOutput()
            videoOutput?.alwaysDiscardsLateVideoFrames = false // TODO: is this necessary?
            videoOutput?.setSampleBufferDelegate(self, queue: videoQueue)

            if let v = videoOutput {
                captureSession.addOutput(v)
            }

            audioOutput = AVCaptureAudioDataOutput()
            audioOutput?.setSampleBufferDelegate(self, queue: videoQueue)

            if let a = audioOutput {
                captureSession.addOutput(a)
            }

            // configure audio session
            let audioSession = AVAudioSession.sharedInstance()
            try audioSession.setCategory(AVAudioSessionCategoryPlayAndRecord)
            try audioSession.setActive(true)

            var micPort: AVAudioSessionPortDescription?

            if let inputs = audioSession.availableInputs {
                for port in inputs {
                    if port.portType == AVAudioSessionPortBuiltInMic {
                        micPort = port
                        break;
                    }
                }
            }

            if let port = micPort, let dataSources = port.dataSources {

                for source in dataSources {
                    if source.orientation == AVAudioSessionOrientationFront {
                        try audioSession.setPreferredInput(port)
                        break
                    }
                }
            }

        } catch {
            print("Failed to configure audio/video capture session")
            throw error
        }
    }

    private func configureAssetWriter() throws {

        prepareVideoFile()

        do {

            assetWriter = try AVAssetWriter(url: outputUrl, fileType: AVFileTypeMPEG4)

            guard let writer = assetWriter else {
                print("Asset writer not created")
                return
            }

            let vidSize = videoSize
            let videoSettings: [String: Any] = [AVVideoCodecKey: AVVideoCodecH264,
                                 AVVideoWidthKey: NSNumber(value: Float(vidSize.width)),
                                 AVVideoHeightKey: NSNumber(value: Float(vidSize.height))]

            videoInput = AVAssetWriterInput(mediaType: AVMediaTypeVideo, outputSettings: videoSettings)
            videoInput?.expectsMediaDataInRealTime = true
            videoInput?.transform = getVideoTransform()

            var channelLayout = AudioChannelLayout()
            memset(&channelLayout, 0, MemoryLayout<AudioChannelLayout>.size);
            channelLayout.mChannelLayoutTag = kAudioChannelLayoutTag_Stereo;

            let audioSettings: [String: Any] = [AVFormatIDKey: kAudioFormatMPEG4AAC,
                                                AVSampleRateKey: 44100,
                                                AVNumberOfChannelsKey: 2]

            audioInput = AVAssetWriterInput(mediaType: AVMediaTypeAudio, outputSettings: audioSettings)
            audioInput?.expectsMediaDataInRealTime = true

            guard let vi = videoInput else {
                print("Video input not configured")
                return
            }

            guard let ai = audioInput else {
                print("Audio input not configured")
                return
            }

            if writer.canAdd(vi) {
                writer.add(vi)
            }

            if writer.canAdd(ai) {
                writer.add(ai)
            }

        } catch {
            print("Failed to configure asset writer")
            throw error
        }
    }

    private func prepareVideoFile() {

        if FileManager.default.fileExists(atPath: outputUrl.path) {

            do {
                try FileManager.default.removeItem(at: outputUrl)
            } catch {
                print("Unable to remove file at URL \(outputUrl)")
            }
        }

        if !FileManager.default.fileExists(atPath: outputDirectory.path) {

            do {
                try FileManager.default.createDirectory(at: outputDirectory, withIntermediateDirectories: true, attributes: nil)
            } catch {
                print("Unable to create directory at URL \(outputDirectory)")
            }
        }
    }

    private func configurePreview() {

        if let previewLayer = AVCaptureVideoPreviewLayer(session: captureSession) {
            previewLayer.frame = previewView.bounds
            previewView.layer.addSublayer(previewLayer)
        }
    }

    private func getVideoSize() -> CGSize {

        if UIDevice.current.orientation == .landscapeLeft || UIDevice.current.orientation == .landscapeRight {

            if videoSize.width > videoSize.height {
                return videoSize
            } else {
                return CGSize(width: videoSize.height, height: videoSize.width)
            }

        } else {

            if videoSize.width < videoSize.height {
                return videoSize
            } else {
                return CGSize(width: videoSize.height, height: videoSize.width)
            }
        }
    }

    private func getVideoTransform() -> CGAffineTransform {

        switch UIDevice.current.orientation {

        case .portraitUpsideDown:
            return CGAffineTransform(rotationAngle: CGFloat((M_PI * -90.0)) / 180.0)

        case .landscapeLeft:
            return CGAffineTransform(rotationAngle: CGFloat((M_PI * -180.0)) / 180.0) // TODO: Add support for front facing camera
//            return CGAffineTransform(rotationAngle: CGFloat((M_PI * 0.0)) / 180.0) // TODO: For front facing camera

        case .landscapeRight:
            return CGAffineTransform(rotationAngle: CGFloat((M_PI * 0.0)) / 180.0) // TODO: Add support for front facing camera
//            return CGAffineTransform(rotationAngle: CGFloat((M_PI * -180.0)) / 180.0) // TODO: For front facing camera

        default:
            return CGAffineTransform(rotationAngle: CGFloat((M_PI * 90.0)) / 180.0)
        }
    }

    //MARK: - Controls

    private func startRecording() {

        videoQueue.async {
            self.captureSession.startRunning()
        }

        isRecording = true
        playStopButton.setTitle("Stop Recording", for: .normal)
        print("Recording did start")
    }

    private func stopRecording() {

        if !isRecording {
            return
        }

        videoQueue.async {

            self.assetWriter?.finishWriting {
                print("Asset writer did finish writing")
                self.isWriting = false
            }

            self.captureSession.stopRunning()
        }

        isRecording = false

        playStopButton.setTitle("Start Recording", for: .normal)
        print("Recording did stop")
    }

    //MARK: - AVCaptureVideoDataOutputSampleBufferDelegate

    func captureOutput(_ captureOutput: AVCaptureOutput!, didOutputSampleBuffer sampleBuffer: CMSampleBuffer!, from connection: AVCaptureConnection!) {

        guard let w = assetWriter else {
            print("Asset writer not configured")
            return
        }

        guard let vo = videoOutput else {
            print("Video output not configured")
            return
        }

        guard let ao = audioOutput else {
            print("Audio output not configured")
            return
        }

        guard let vi = videoInput else {
            print("Video input not configured")
            return
        }

        guard let ai = audioInput else {
            print("Audio input not configured")
            return
        }

        let st = CMSampleBufferGetPresentationTimeStamp(sampleBuffer)

        print("Writer status \(w.status.rawValue)")

        if let e = w.error {
            print("Writer error \(e)")
            stopRecording()
            return
        }

        switch w.status {

        case .unknown:

            if !isWriting {
                isWriting = true
                w.startWriting()
                w.startSession(atSourceTime: st)
            }

            return

        case .completed:
            print("Video writing completed")
            return

        case .cancelled:
            print("Video writing cancelled")
            return

        case .failed:
            print("Video writing failed")
            return

        default:
            print("Video is writing")
        }

        if vo == captureOutput {

            if !vi.append(sampleBuffer) {
                print("Unable to write to video buffer")
            }

        } else if ao == captureOutput {

            if !ai.append(sampleBuffer) {
                print("Unable to write to audio buffer")
            }
        }
    }

    //MARK: Helpers

    private func getDocumentsDirectory() -> URL {
        let paths = FileManager.default.urls(for: .documentDirectory, in: .userDomainMask)
        let documentsDirectory = paths[0]
        return documentsDirectory
    }

    //MARK: Actions

    @IBAction func startStopTapped(sender: AnyObject) {

        if isRecording {
            stopRecording()
        } else {
            startRecording()
        }
    }
}

最佳答案

视频方向由 AVAssetWriterInput.transform 处理,看起来 getVideoTransform() 实现不正确 - CGAffineTransform 需要旋转角度以弧度为单位,因此需要更改为如下所示:

private func getVideoTransform() -> CGAffineTransform {
    switch UIDevice.current.orientation {
        case .portrait:
            return .identity
        case .portraitUpsideDown:
            return CGAffineTransform(rotationAngle: .pi)
        case .landscapeLeft:
            return CGAffineTransform(rotationAngle: .pi/2)
        case .landscapeRight:
            return CGAffineTransform(rotationAngle: -.pi/2)
        default:
            return .identity
        }
    }

来自 Apple 技术问答: https://developer.apple.com/library/archive/qa/qa1744/_index.html

If you are using an AVAssetWriter object to write a movie file, you can use the transform property of the associated AVAssetWriterInput to specify the output file orientation. This will write a display transform property into the output file as the preferred transformation of the visual media data for display purposes. See the AVAssetWriterInput.h interface file for the details.

关于iOS 如何在使用 AVAssetWriter 捕获视频时正确处理方向,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/42554122/

相关文章:

ios - 在 UITableView 的单元格中选择特定的字符串值

ios - NSSplitView 建议的分割分隔线没有放置在正确的位置?

regex - 如何在swift中通过正则表达式过滤字典键

c# - 尝试使用 ffmpeg 加入两个 mp4 视频文件但出现错误

c# - 如何从视频中抓取一帧?

ios - iTunes Connect - 无法为预发布应用邀请 "internal testers"

ios - “可用更新”弹出窗口不显示

sql - 在 Swift 中,使用带有撇号的 SQL 查询

ios - 带 bo​​ol 的空合并运算符

python - 如何在Python中获取在线视频文件的视频时长?