Create a StoryTeller in Swift with Firebase ML Kit Text Recognition OCR | Lil ‘Bits | https://www.youtube.com/watch?v=kpzyFZYI1PQ

StoryTellerViewController.swift

import UIKit
import AVFoundation
import Firebase

class StoryTellerViewController: UIViewController {

    private lazy var previewLayer = AVCaptureVideoPreviewLayer(session: captureSession)
    private lazy var captureSession: AVCaptureSession = {
        let captureSession = AVCaptureSession()
        guard let captureDevice = AVCaptureDevice.default(.builtInWideAngleCamera, for: .video, position: .back),
            let input = try? AVCaptureDeviceInput(device: captureDevice) else {
                return captureSession
        }
        captureSession.addInput(input)
        let output = AVCaptureVideoDataOutput()
        output.setSampleBufferDelegate(self, queue: DispatchQueue(label: "SampleBuffer"))
        captureSession.addOutput(output)
        return captureSession
    }()

    let speechSynthesizer = AVSpeechSynthesizer()
    lazy var vision = Vision.vision()
    var textDetector: VisionTextDetector?
    private var textsToRead = [String]()

    override func viewDidLoad() {
        super.viewDidLoad()
        speechSynthesizer.delegate = self
        view.layer.addSublayer(previewLayer)
        try? setPlaybackAudioSession()
    }

    override func viewWillAppear(_ animated: Bool) {
        super.viewWillAppear(animated)
        previewLayer.frame = view.frame
        checkCameraAccess()
    }

    func checkCameraAccess() {
        switch AVCaptureDevice.authorizationStatus(for: .video) {
            case .authorized: captureSession.startRunning()
            default: requestCameraAccess()
        }
    }

    func requestCameraAccess() {
        AVCaptureDevice.requestAccess(for: .video) { _ in
            self.checkCameraAccess()
        }
    }

    func setPlaybackAudioSession() throws {
        let audioSession = AVAudioSession.sharedInstance()
        try audioSession.setCategory(AVAudioSessionCategoryPlayback)
        try audioSession.setMode(AVAudioSessionModeMeasurement)
        try audioSession.setActive(true, with: .notifyOthersOnDeactivation)
    }

    func read(_ texts: [String]) {
        textsToRead.append(contentsOf: texts)
        read(textsToRead.removeFirst())
    }

    func read(_ text: String) {
        let speechUtterance = AVSpeechUtterance(string: text)
        let dominantLanguage = NSLinguisticTagger.dominantLanguage(for: text)
        speechUtterance.voice = AVSpeechSynthesisVoice(language: dominantLanguage)
        speechSynthesizer.speak(speechUtterance)
    }

}

extension StoryTellerViewController: AVSpeechSynthesizerDelegate {

    func speechSynthesizer(_ synthesizer: AVSpeechSynthesizer, didFinish utterance: AVSpeechUtterance) {
        if textsToRead.isEmpty {
            textDetector = nil
            captureSession.startRunning()
        } else {
            read(textsToRead.removeFirst())
        }
    }

}

extension StoryTellerViewController: AVCaptureVideoDataOutputSampleBufferDelegate {

    func captureOutput(_ output: AVCaptureOutput, didOutput sampleBuffer: CMSampleBuffer, from connection: AVCaptureConnection) {
        if textDetector == nil {
            let metadata = VisionImageMetadata()
            metadata.orientation = .rightTop
            let visionImage = VisionImage(buffer: sampleBuffer)
            visionImage.metadata = metadata
            textDetector = vision.textDetector()
            textDetector?.detect(in: visionImage, completion: handleDetect)
        }
    }

    func handleDetect(features: [VisionText]?, error: Error?) {
        guard error == nil, let features = features, !features.isEmpty else {
            textDetector = nil
            return
        }
        captureSession.stopRunning()
        read(features.compactMap { $0.text })
    }

}