iOS---語音轉文字

weixin_34050427發表於2018-05-26

iOS中實現語音轉文字,除了一些第三方,如科大訊飛語音、百度語音等第三方的外(這種第三方的在其官方都有詳細的教程,這裡就不在敘述了。特點是:百度語音支援離線並免費。訊飛的也支援離線,雖然識別率很高也很精準,但是收費),蘋果官方也推出了自己的一套識別標準。

蘋果官方的文字轉語音支援iOS10之後

文字轉語音的需求:

1.建立一個按鈕用來控制啟動、關閉錄音

 2.建立一個語音控制器,一個語音識別器 。一個語音工作管理員,

 3.建立顯示轉換好的文字的控制元件

步驟:

 iOS語音轉文字:1.需要新增Speech.framework庫    2.匯入標頭檔案#import


 3.在info.plist檔案裡新增了兩個鍵值Privacy - Microphone Usage Description、Privacy - Speech Recognition Usage Description

具體程式碼如下:

#import "ViewController.h"

#import<Speech/Speech.h>

#import<AVFoundation/AVFoundation.h>

@interface ViewController ()

@property (nonatomic,strong)UIButton *swicthBut;

@property (nonatomic,strong)UILabel *labText;

@property(nonatomic,strong)SFSpeechRecognizer*speechRecognizer;//語音識別器

@property (nonatomic,strong) SFSpeechAudioBufferRecognitionRequest *recognitionRequest;//語音識別請求

@property (nonatomic, strong) SFSpeechRecognitionTask *recognitionTask;//語音工作管理員

@property (nonatomic,strong) AVAudioEngine *audioEngine;//語音控制器

@end

@implementation ViewController

- (void)viewDidLoad {

    [super viewDidLoad];

    //傳送語音認證請求(首先要判斷裝置是否支援語音識別功能)

    [SFSpeechRecognizer requestAuthorization:^(SFSpeechRecognizerAuthorizationStatus status) {

        boolisButtonEnabled =false;

        switch(status) {

            case SFSpeechRecognizerAuthorizationStatusAuthorized:

                isButtonEnabled =true;

                NSLog(@"可以語音識別");

                break;

            case SFSpeechRecognizerAuthorizationStatusDenied:

                isButtonEnabled =false;

                NSLog(@"使用者未授權使用語音識別");

                break;

            case SFSpeechRecognizerAuthorizationStatusRestricted:

                isButtonEnabled =false;

                NSLog(@"語音識別在這臺裝置上受到限制");

                break;

            case SFSpeechRecognizerAuthorizationStatusNotDetermined:

                isButtonEnabled =false;

                NSLog(@"語音識別未授權");

                break;

            default:

                break;

        }


    }];



    // Do any additional setup after loading the view, typically from a nib.

    self.title=@"測試";

    self.view.backgroundColor = [UIColor whiteColor];

    [self.viewaddSubview:self.swicthBut];

    [self.viewaddSubview:self.labText];



}

- (void)speechRecognizer:(SFSpeechRecognizer*)speechRecognizer availabilityDidChange:(BOOL)available{

    if(available) {

        self.swicthBut.enabled=YES;

          [self.swicthBut setTitle:@"開始錄音" forState:UIControlStateNormal];

    }else{


        self.swicthBut.enabled=NO;

          [self.swicthBut setTitle:@"語音識別不可用" forState:UIControlStateNormal];

    }

}

#pragma mark----語音識別

- (SFSpeechRecognizer*)speechRecognizer{


    if (!_speechRecognizer) {

        NSLocale *cale = [[NSLocale alloc]initWithLocaleIdentifier:@"zh-CN"];

       _speechRecognizer= [[SFSpeechRecognizeralloc]initWithLocale:cale];


        //設定代理

        _speechRecognizer.delegate = self;

    }

    return _speechRecognizer;

}

#pragma mark---停止錄音

- (void)endRecording{

    [self.audioEngine stop];

    if (_recognitionRequest) {

        [_recognitionRequest endAudio];

    }


    if (_recognitionTask) {

        [_recognitionTask cancel];

        _recognitionTask = nil;

    }

    self.swicthBut.enabled=NO;

}

#pragma  mark---開始錄音

- (void)startRecording{

    if (self.recognitionTask) {

        [self.recognitionTask cancel];

        self.recognitionTask = nil;

    }

    AVAudioSession *audioSession = [AVAudioSession sharedInstance];

    NSError*error;

    bool  audioBool = [audioSessionsetCategory:AVAudioSessionCategoryRecorderror:&error];

    NSParameterAssert(!error);

    bool  audioBool1= [audioSessionsetMode:AVAudioSessionModeMeasurementerror:&error];

      NSParameterAssert(!error);

    bool  audioBool2= [audioSessionsetActive:true withOptions:AVAudioSessionSetActiveOptionNotifyOthersOnDeactivation error:&error];

      NSParameterAssert(!error);

    if(audioBool || audioBool1||  audioBool2) {

        NSLog(@"可以使用");

    }else{

        NSLog(@"這裡說明有的功能不支援");

    }

    self.recognitionRequest = [[SFSpeechAudioBufferRecognitionRequest alloc]init];

    AVAudioInputNode *inputNode = self.audioEngine.inputNode;

    NSAssert(inputNode,@"錄入裝置沒有準備好");

    NSAssert(self.recognitionRequest, @"請求初始化失敗");


    self.recognitionRequest.shouldReportPartialResults = true;

     __weaktypeof(self) weakSelf =self;


    //開始識別任務

    self.recognitionTask = [self.speechRecognizer recognitionTaskWithRequest:self.recognitionRequest resultHandler:^(SFSpeechRecognitionResult * _Nullable result, NSError * _Nullable error) {

        __strongtypeof(weakSelf) strongSelf = weakSelf;

        boolisFinal =false;

        if(result) {

            strongSelf.labText.text= [[resultbestTranscription]formattedString];//語音轉文字

            isFinal = [resultisFinal];

        }

        if(error || isFinal) {

            [strongSelf.audioEnginestop];

            [inputNoderemoveTapOnBus:0];

            strongSelf.recognitionRequest=nil;

            strongSelf.recognitionTask=nil;

            [strongSelf.swicthButsetTitle:@"開始錄音"forState:UIControlStateNormal];

            strongSelf.swicthBut.enabled=true;

        }

    }];

    AVAudioFormat*recordingFormat = [inputNodeoutputFormatForBus:0];

    //在新增tap之前先移除上一個  不然有可能報"Terminating app due to uncaught exception 'com.apple.coreaudio.avfaudio',"之類的錯誤

    [inputNoderemoveTapOnBus:0];

    [inputNodeinstallTapOnBus:0bufferSize:1024format:recordingFormatblock:^(AVAudioPCMBuffer*_Nonnullbuffer,AVAudioTime*_Nonnullwhen) {

         __strongtypeof(weakSelf) strongSelf = weakSelf;

        if(strongSelf.recognitionRequest) {

            [strongSelf.recognitionRequestappendAudioPCMBuffer:buffer];

        }


    }];

    [self.audioEngine prepare];

    boolaudioEngineBool = [self.audioEnginestartAndReturnError:&error];

     NSParameterAssert(!error);

    self.labText.text = @"正在錄音。。。";

    NSLog(@"%d",audioEngineBool);



}

#pragma mark---建立錄音引擎

- (AVAudioEngine*)audioEngine{


    if (!_audioEngine) {

        _audioEngine= [[AVAudioEnginealloc]init];

    }

    return _audioEngine;

}

#pragma mark----顯示控制元件

- (UILabel*)labText{


    if(!_labText) {

        _labText= [[UILabelalloc]init];

        _labText.frame = CGRectMake(0, 140, [UIScreen mainScreen].bounds.size.width, 50);

        _labText.font= [UIFontsystemFontOfSize:13.0f];

        _labText.numberOfLines = 0;

        _labText.textAlignment = NSTextAlignmentCenter;

        _labText.textColor = [UIColor yellowColor];

    }

    return _labText;

}

#pragma mark----開關

- (UIButton*)swicthBut{


    if (!_swicthBut) {

        _swicthBut= [[UIButtonalloc]init];

        _swicthBut.frame=CGRectMake(50,100,80,30);

        _swicthBut.titleLabel.textAlignment = NSTextAlignmentCenter;

        [_swicthBut setTitleColor:[UIColor orangeColor] forState:UIControlStateNormal];

        [_swicthBut addTarget:self action:@selector(switchOn:) forControlEvents:UIControlEventTouchUpInside];

        [_swicthBut setTitle:@"開始錄音" forState:UIControlStateNormal];

    }

    return _swicthBut;

}

- (void)switchOn:(id)sender{


    if([self.audioEngineisRunning]) {

        [self endRecording];

       [_swicthBut setTitle:@"開始錄音" forState:UIControlStateNormal];

    }else{

        [self startRecording];

         [_swicthBut setTitle:@"關閉" forState:UIControlStateNormal];

    }

}

#pragma mark---識別本地音訊檔案

- (void)recognizeLocalAudioFile:(UIButton*)sender {

    NSLocale *local =[[NSLocale alloc] initWithLocaleIdentifier:@"zh_CN"];

    SFSpeechRecognizer *localRecognizer =[[SFSpeechRecognizer alloc] initWithLocale:local];

    NSURL *url =[[NSBundle mainBundle] URLForResource:@"錄音.m4a" withExtension:nil];

    if(!url)return;

    SFSpeechURLRecognitionRequest *res =[[SFSpeechURLRecognitionRequest alloc] initWithURL:url];

    __weak typeof(self) weakSelf = self;

    [localRecognizerrecognitionTaskWithRequest:resresultHandler:^(SFSpeechRecognitionResult*_Nullableresult,NSError*_Nullableerror) {

        if(error) {

            NSLog(@"語音識別解析失敗,%@",error);

        }

        else

        {

            weakSelf.labText.text = result.bestTranscription.formattedString;

        }

    }];

}

- (void)didReceiveMemoryWarning {

    [super didReceiveMemoryWarning];

    // Dispose of any resources that can be recreated.

}

@end

相關文章