ios开发科大讯飞的语音识别

xiaoxiao2024-04-19 5

1、新建工程不说了

2、导入库：xcode工程下没有Frameworks目录了，可以自己newgroup一个出来，然后在重命名一下，注意大小写，不然导入后不在该目录下，当然没有强迫症的可以忽略，直接导入就好了，新建目录是结构看起来舒服点。。。

第三方库，即iFly的库，到科大讯飞上申请，然后下载sdk，把lib目录下的iflyMSC.framework拷贝到工程目录下，一定要拷贝，不然导进去会出现错误，上网搜了一下，发现好多了都进过这个坑。。。还有就是一些依赖库，直接add进去就好了

3、添加桥接文件：

命名“工程名-Bridging-Header.h”，然后把import的直接copy进去就好了

// // testiFly-Bridging-Header.h // testiFly // // Created by fengdelei on 16/8/12. // Copyright © 2016年 fengdelei. All rights reserved. // #ifndef testiFly_Bridging_Header_h #define testiFly_Bridging_Header_h #import "iflyMSC/IFlySpeechRecognizerDelegate.h" #import "iflyMSC/IFlySpeechRecognizer.h" #import "iflyMSC/IFlyRecognizerViewDelegate.h" #import "iflyMSC/IFlyRecognizerView.h" #import "iflyMSC/IFlyContact.h" #import "iflyMSC/IFlyUserWords.h" #import "iflyMSC/IFlyDataUploader.h" #import "iflyMSC/IFlySpeechSynthesizerDelegate.h" #import "iflyMSC/IFlySpeechSynthesizer.h" #import "iflyMSC/IFlySpeechUtility.h" #import "iflyMSC/IFlySpeechConstant.h" #import "iflyMSC/IFlySpeechError.h" #import "iflyMSC/IFlySpeechUnderstander.h" #import "iflyMSC/IFlyTextUnderstander.h" #import "iflyMSC/IFlySetting.h" #endif

然后把桥接文件配置一下，Build Settings->Swift Compiler-〉Objective-C Bridging Header，空白处点击，然后把.h文件直接拖进去就好了，直接出来路径的

4、其他一点配置：

iFly还不运行bitcode，设置为no

ios9中新增ats我，主要是用到的http请求都转向tls1.2协议进行传输，强制使用了HTTPS协议进行传输，把info.plist添加如图属性就好了：

好了，环境已经配置好了，代码：

</pre><pre name="code" class="plain">import UIKit class HomeViewController: UIViewController, IFlyRecognizerViewDelegate { @IBOutlet weak var voice_img: UIImageView! var iflyRecognizerView:IFlyRecognizerView! var isRecongnizer = false override func viewDidLoad() { super.viewDidLoad() var initString:String! initString = "appid=XXXXXXXX" IFlySpeechUtility.createUtility(initString) self.iflyRecognizerView = IFlyRecognizerView.init(center: self.view.center)as IFlyRecognizerView self.iflyRecognizerView.delegate = self self.iflyRecognizerView.setParameter("iat", forKey: IFlySpeechConstant.IFLY_DOMAIN()) self.iflyRecognizerView.setParameter("16000", forKey: IFlySpeechConstant.SAMPLE_RATE()) // | result_type | 返回结果的数据格式 plain,只支持plain self.iflyRecognizerView.setParameter("plain", forKey: IFlySpeechConstant.RESULT_TYPE()) // Do any additional setup after loading the view, typically from a nib. self.view.backgroundColor = UIColor.blackColor(); voice_img.image = UIImage.init(named: "app_voice.png"); voice_img.userInteractionEnabled = true let voice_tap = UITapGestureRecognizer(target: self , action: #selector(HomeViewController.startVoiceBtn)) voice_img.addGestureRecognizer(voice_tap) } override func didReceiveMemoryWarning() { super.didReceiveMemoryWarning() // Dispose of any resources that can be recreated. } func startVoiceBtn() { print("开始识别") iflyRecognizerView.start() } func onResult(results: [AnyObject]!,isLast: Bool) { var resultText : string ="" var resultStr : String = "" if results != nil { let resultDic : Dictionary<String, String> = results[0] as! Dictionary<String, String> for key in resultDic.keys { resultStr += key } } if resultText != "" { if (resultText as NSString).substringWithRange(NSMakeRange( resultText.characters.count - 1, 1)) != "," { resultText += "," } } resultText += resultStr //textView.text = resultText print(resultText) if isRecongnizer { iflyRecognizerView.start() } else { iflyRecognizerView.cancel() if resultText != "" { resultText = (resultText as NSString).substringWithRange(NSMakeRange( 0, resultText.characters.count - 1)) + "。" //textView.text = resultText print(resultText) } } } func onError(error: IFlySpeechError!) { print("识别出错：\(error.errorCode)") } }

代码是从工程里扣出来的，删掉了其他一部分不相关的，应该没有误删语音识别的相关东西，自己试一下吧，等有空我，整理一份上传一下。appid注意要替换一下！！！好啦，可以看log里的结果了，也可以自己写个textview，把结果打到界面上去。我这里试了几次，结果不是很好，同事试的骂人的话，倒是挺准的，不知道什么情况。说的你好，一直都是“你。”普通话这么差吗？

参考：http://www.zhimengzhe.com/bianchengjiaocheng/swift/83732.html

转载请注明原文地址: https://ju.6miu.com/read-1288144.html

最新回复(0)