PHPAPI语音识别怎么集成

wen PHP项目 2026-07-06 8

本文目录导读：

PHPAPI语音识别怎么集成

主流语音识别API选择
通用PHP语音识别类
音频预处理
完整集成示例
依赖安装
注意事项

我来详细介绍PHP集成语音识别API的方法：

主流语音识别API选择

百度语音识别

<?php
// 使用百度语音识别API
require_once 'AipSpeech.php';
// 配置AppID、API Key、Secret Key
const APP_ID = '你的AppID';
const API_KEY = '你的API Key';
const SECRET_KEY = '你的Secret Key';
$client = new AipSpeech(APP_ID, API_KEY, SECRET_KEY);
// 读取音频文件
$audio = file_get_contents('/path/to/audio.pcm');
// 调用语音识别
$result = $client->asr($audio, 'pcm', 16000, [
    'dev_pid' => 1536, // 普通话(支持简单英文)
]);
print_r($result);
?>

阿里云语音识别

<?php
// 使用阿里云语音识别
use AlibabaCloud\Client\AlibabaCloud;
use AlibabaCloud\Client\Exception\ClientException;
use AlibabaCloud\Client\Exception\ServerException;
// 配置AK
AlibabaCloud::accessKeyClient('your-access-key-id', 'your-access-key-secret')
    ->regionId('cn-shanghai')
    ->asDefaultClient();
try {
    $result = AlibabaCloud::nlsFiletrans()
        ->v20180817()
        ->submitTask()
        ->withFileUrl('https://your-bucket.oss-cn-shanghai.aliyuncs.com/audio.wav')
        ->withCallbackUrl('http://your-callback-url.com/callback')
        ->withAppkey('your-appkey')
        ->request();
    print_r($result->toArray());
} catch (ClientException $e) {
    echo $e->getErrorMessage() . PHP_EOL;
} catch (ServerException $e) {
    echo $e->getErrorMessage() . PHP_EOL;
}
?>

腾讯云语音识别

<?php
// 使用腾讯云语音识别
require_once './vendor/autoload.php';
use TencentCloud\Common\Credential;
use TencentCloud\Common\Profile\ClientProfile;
use TencentCloud\Common\Profile\HttpProfile;
use TencentCloud\Asr\V20190614\AsrClient;
use TencentCloud\Asr\V20190614\Models\CreateRecTaskRequest;
$cred = new Credential("your-secret-id", "your-secret-key");
$httpProfile = new HttpProfile();
$httpProfile->setEndpoint("asr.tencentcloudapi.com");
$clientProfile = new ClientProfile();
$clientProfile->setHttpProfile($httpProfile);
$client = new AsrClient($cred, "ap-guangzhou", $clientProfile);
$req = new CreateRecTaskRequest();
$params = array(
    "EngineModelType" => "16k_0",
    "ChannelNum" => 1,
    "ResTextFormat" => 0,
    "SourceType" => 0,
    "Url" => "https://your-audio-url.com/audio.wav"
);
$req->fromJsonString(json_encode($params));
$resp = $client->CreateRecTask($req);
print_r(json_decode($resp->toJsonString(), true));
?>

通用PHP语音识别类

<?php
class SpeechRecognition {
    private $apiKey;
    private $apiUrl;
    public function __construct($apiKey, $apiUrl) {
        $this->apiKey = $apiKey;
        $this->apiUrl = $apiUrl;
    }
    /**
     * 识别音频文件
     * @param string $audioFilePath 音频文件路径
     * @param string $format 音频格式
     * @return array
     */
    public function recognize($audioFilePath, $format = 'wav') {
        if (!file_exists($audioFilePath)) {
            return ['error' => '音频文件不存在'];
        }
        // 读取音频文件
        $audioData = file_get_contents($audioFilePath);
        // 转换为base64
        $base64Audio = base64_encode($audioData);
        // 发送请求
        $response = $this->sendRequest($base64Audio, $format);
        return json_decode($response, true);
    }
    /**
     * 发送请求到语音识别API
     */
    private function sendRequest($audioData, $format) {
        $ch = curl_init();
        curl_setopt($ch, CURLOPT_URL, $this->apiUrl);
        curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
        curl_setopt($ch, CURLOPT_POST, 1);
        curl_setopt($ch, CURLOPT_HTTPHEADER, [
            'Content-Type: application/json',
            'Authorization: Bearer ' . $this->apiKey
        ]);
        $requestData = [
            'audio' => [
                'content' => $audioData,
                'format' => $format
            ],
            'config' => [
                'language' => 'zh-CN', // 语言
                'sample_rate' => 16000  // 采样率
            ]
        ];
        curl_setopt($ch, CURLOPT_POSTFIELDS, json_encode($requestData));
        $response = curl_exec($ch);
        $httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);
        if ($httpCode !== 200) {
            return json_encode(['error' => 'API请求失败', 'http_code' => $httpCode]);
        }
        curl_close($ch);
        return $response;
    }
    /**
     * 实时语音识别（WebSocket）
     */
    public function realTimeRecognize($audioStream, $callback) {
        // WebSocket实现实时识别
        // 这里需要根据具体的API实现
        throw new Exception('实时识别需要具体API支持');
    }
}
// 使用示例
$recognizer = new SpeechRecognition(
    'your-api-key',
    'https://api.example.com/v1/speech/recognize'
);
$result = $recognizer->recognize('/path/to/audio.wav');
if (!isset($result['error'])) {
    echo "识别结果: " . $result['text'];
} else {
    echo "错误: " . $result['error'];
}
?>

音频预处理

<?php
/**
 * 音频预处理类
 */
class AudioPreprocessor {
    /**
     * 转换音频格式
     */
    public static function convertFormat($inputFile, $outputFile, $targetFormat = 'wav') {
        $command = "ffmpeg -i {$inputFile} -acodec pcm_s16le -ar 16000 -ac 1 {$outputFile}";
        exec($command, $output, $returnCode);
        return $returnCode === 0;
    }
    /**
     * 剪切音频
     */
    public static function trimAudio($inputFile, $outputFile, $startTime, $duration) {
        $command = "ffmpeg -i {$inputFile} -ss {$startTime} -t {$duration} -c copy {$outputFile}";
        exec($command, $output, $returnCode);
        return $returnCode === 0;
    }
    /**
     * 检测音频是否有效
     */
    public static function validateAudio($audioFile) {
        $command = "ffprobe -v quiet -print_format json -show_format " . escapeshellarg($audioFile);
        exec($command, $output, $returnCode);
        if ($returnCode !== 0) {
            return false;
        }
        $info = json_decode(implode('', $output), true);
        $format = $info['format'];
        // 检查格式和大小
        if (!isset($format['duration']) || $format['duration'] > 60) {
            return false; // 音频超过60秒
        }
        if (!isset($format['size']) || $format['size'] > 10 * 1024 * 1024) {
            return false; // 文件超过10MB
        }
        return true;
    }
}
?>

完整集成示例

<?php
// index.php - 语音识别示例页面
require_once 'SpeechRecognition.php';
require_once 'AudioPreprocessor.php';
// 处理上传的音频文件
if ($_SERVER['REQUEST_METHOD'] === 'POST' && isset($_FILES['audio'])) {
    $uploadDir = 'uploads/';
    $convertedDir = 'converted/';
    // 创建目录
    if (!file_exists($uploadDir)) mkdir($uploadDir, 0777, true);
    if (!file_exists($convertedDir)) mkdir($convertedDir, 0777, true);
    $uploadFile = $uploadDir . basename($_FILES['audio']['name']);
    $convertedFile = $convertedDir . 'converted_' . time() . '.wav';
    // 上传文件
    if (move_uploaded_file($_FILES['audio']['tmp_name'], $uploadFile)) {
        // 预处理音频
        if (AudioPreprocessor::validateAudio($uploadFile)) {
            // 转换格式
            AudioPreprocessor::convertFormat($uploadFile, $convertedFile, 'wav');
            // 初始化识别器
            $recognizer = new SpeechRecognition(
                'your-api-key',
                'https://api.speech-recognition.com/v1/recognize'
            );
            // 执行识别
            $result = $recognizer->recognize($convertedFile);
            // 显示结果
            if (isset($result['text'])) {
                echo "识别结果: " . htmlspecialchars($result['text']);
            } else {
                echo "识别失败: " . json_encode($result);
            }
            // 清理临时文件
            unlink($uploadFile);
            unlink($convertedFile);
        } else {
            echo "音频文件无效";
        }
    } else {
        echo "文件上传失败";
    }
}
?>
<!DOCTYPE html>
<html>
<head>语音识别演示</title>
</head>
<body>
    <h1>语音识别演示</h1>
    <form method="post" enctype="multipart/form-data">
        <input type="file" name="audio" accept="audio/*" required>
        <button type="submit">识别语音</button>
    </form>
    <!-- 或使用麦克风在线录音 -->
    <button id="startBtn">开始录音</button>
    <button id="stopBtn" disabled>停止录音</button>
    <script>
    let mediaRecorder;
    let audioChunks = [];
    document.getElementById('startBtn').onclick = async function() {
        const stream = await navigator.mediaDevices.getUserMedia({ audio: true });
        mediaRecorder = new MediaRecorder(stream);
        mediaRecorder.ondataavailable = event => {
            audioChunks.push(event.data);
        };
        mediaRecorder.onstop = function() {
            const audioBlob = new Blob(audioChunks, { type: 'audio/wav' });
            const formData = new FormData();
            formData.append('audio', audioBlob, 'recording.wav');
            fetch('', {
                method: 'POST',
                body: formData
            }).then(response => response.text())
              .then(result => alert(result));
        };
        mediaRecorder.start();
        document.getElementById('startBtn').disabled = true;
        document.getElementById('stopBtn').disabled = false;
    };
    document.getElementById('stopBtn').onclick = function() {
        mediaRecorder.stop();
        document.getElementById('startBtn').disabled = false;
        document.getElementById('stopBtn').disabled = true;
    };
    </script>
</body>
</html>

依赖安装

# 安装必要的PHP扩展
sudo apt-get install php-curl php-json
# 安装FFmpeg用于音频处理
sudo apt-get install ffmpeg
# 使用Composer安装SDK（以阿里云为例）
composer require alibabacloud/client
# 或其他云服务商SDK
composer require qcloud/cos-sdk-v5

注意事项

API Key安全：不要在代码中硬编码，使用环境变量
音频格式：通常需要16kHz、16bit、单声道PCM或WAV格式
文件大小限制：多数API有文件大小限制（通常10-20MB）
错误处理：完善的错误处理和重试机制
成本控制：注意API调用次数限制和费用

建议优先选择国内云服务商的API，延迟更小,支持更全面。