一、接口介绍
基于文字识别与文本翻译技术，满足用户翻译图片文字的需求。只需要通过调用图片翻译API，传入图片的Base64编码，指定源语言与目标语言，通过POST请求方式，就可以识别图片中的文字并进行翻译。
协议须知：
二、申请AppID、密钥
1.登录有道智云AI开放平台，进入控制台
2.在应用总览中创建应用，填写相关内容
3.获取应用ID、密钥
三、在Unity中应用
1.定义请求数据结构
根据官方文档中接口调用参数说明定义对应的数据结构：
/// <summary>
/// 图片翻译请求数据结构
/// </summary>
[Serializable]
public class OcrTransRequest
{
/// <summary>
/// 文件上传类型
/// 目前支持Base64 请置该字段为1
/// </summary>
public string type;
/// <summary>
/// 源语言
/// </summary>
public string from;
/// <summary>
/// 目标语言
/// </summary>
public string to;
/// <summary>
/// 应用ID
/// </summary>
public string appKey;
/// <summary>
/// UUID 唯一通用识别码
/// </summary>
public string salt;
/// <summary>
/// 签名 MD5
/// 应用ID q salt 应用密钥
/// </summary>
public string sign;
/// <summary>
/// 翻译结果音频格式 支持mp3
/// </summary>
public string ext;
/// <summary>
/// 要识别的图片
/// type为1时，图片的base64编码
/// </summary>
public string q;
/// <summary>
/// 服务器响应类型 目前只支持json
/// </summary>
public string docType;
/// <summary>
/// 是否需要服务端返回渲染的图片 0否 1是
/// </summary>
public string render;
public OcrTransRequest(string from, string to, string appKey, string appSecret, string q)
{
type = "1";
this.from = from;
this.to = to;
this.appKey = appKey;
salt = DateTime.Now.Millisecond.ToString();
//拼接应用ID q salt 应用密钥
string signStr = appKey q salt appSecret;
//获取字节数据
byte[] inputBytes = Encoding.UTF8.GetBytes(signStr);
//获取哈希数据
byte[] hashBytes = new MD5CryptoServiceProvider().ComputeHash(inputBytes);
sign = BitConverter.ToString(hashBytes).Replace("-", "");
this.q = HttpUtility.UrlEncode(q);
ext = "mp3";
docType = "json";
render = "0";
}
public override string ToString()
{
return string.Format("from={0}&to={1}&type={2}&q={3}&appKey={4}&salt={5}&sign={6}", from, to, type, q, appKey, salt, sign);
}
}2.定义响应数据结构
根据官方文档中接口返回结果参数说明定义对应的数据结构：
/// <summary>
/// 图片翻译响应数据结构
/// </summary>
[Serializable]
public class OcrTransResponse
{
/// <summary>
/// 图片所对应的方向
/// </summary>
public string orientation;
/// <summary>
/// 图片中的语言
/// </summary>
public string lanFrom;
/// <summary>
/// 图片的倾斜角度
/// </summary>
public string textAngle;
/// <summary>
/// 错误码
/// </summary>
public string errorCode;
/// <summary>
/// 目标语言
/// </summary>
public string lanTo;
/// <summary>
/// 图片翻译的具体内容
/// </summary>
public ResRegion[] resRegions;
}
/// <summary>
/// 图片翻译的具体内容
/// </summary>
[Serializable]
public class ResRegion
{
/// <summary>
/// 区域范围 包含四个值
/// 左上角的x值 左上角的y值 区域宽度 区域高度
/// </summary>
public string boundingBox;
/// <summary>
/// 行数
/// </summary>
public int linesCount;
/// <summary>
/// 行高
/// </summary>
public int lineheight;
/// <summary>
/// 该区域的原文
/// </summary>
public string context;
/// <summary>
/// 行间距
/// </summary>
public int lineSpace;
/// <summary>
/// 翻译结果
/// </summary>
public string tranContent;
/// <summary>
/// 当render=1即返回渲染图片，返回结果lines
/// </summary>
public string lines;
/// <summary>
/// 图片颜色
/// </summary>
public string color;
/// <summary>
/// 行识别结果
/// </summary>
public string text;
/// <summary>
/// 识别的字的结果
/// </summary>
public string word;
/// <summary>
/// 文字的高度
/// </summary>
public int textHeight;
}3.封装调用函数
代码语言：javascript复制/// <summary>
/// 图片翻译
/// </summary>
public class OcrTrans
{
//应用ID和密钥 在有道智云AI开放平台创建应用获取
private static readonly string appid = "";
private static readonly string secret = "";
public static OcrTransResponse SendRequest(string from, string to, string picbase64)
{
string result = "";
string url = "https://openapi.youdao.com/ocrtransapi";
HttpWebRequest request = (HttpWebRequest)WebRequest.Create(url);
request.Method = "POST";
request.ContentType = "application/x-www-form-urlencoded";
OcrTransRequest otr = new OcrTransRequest(from, to, appid, secret, picbase64);
byte[] data = Encoding.UTF8.GetBytes(otr.ToString());
request.ContentLength = data.Length;
using (Stream reqStream = request.GetRequestStream())
{
reqStream.Write(data, 0, data.Length);
reqStream.Close();
}
HttpWebResponse resp = (HttpWebResponse)request.GetResponse();
Stream stream = resp.GetResponseStream();
using (StreamReader reader = new StreamReader(stream, Encoding.UTF8))
{
result = reader.ReadToEnd();
}
Debug.Log(result);
OcrTransResponse response = JsonUtility.FromJson<OcrTransResponse>(result);
return response;
}
}4.测试
using System;
using System.IO;
using UnityEngine;
public class OcrTransExample : MonoBehaviour
{
private void Start()
{
byte[] bytes = File.ReadAllBytes(Application.dataPath "/test.png");
string base64 = Convert.ToBase64String(bytes);
OcrTrans.SendRequest("en", "zh-CHS", base64);
}
}
各语言对应代码：
