使用百度智能云api进行图片识别文字以及pdf识别文字

最新推荐文章于 2024-06-08 20:51:13 发布

卫庄大人℡

最新推荐文章于 2024-06-08 20:51:13 发布

阅读量1.1k

点赞数 8

文章标签： java ocr

本文链接：https://blog.csdn.net/qq_45621643/article/details/136708781

版权

1.登录注册智能百度云并进行实名认证

1.1 个人认证每月有一千次的调用机会

1.2 进入控制台查看相关信息

1.2.1 控制台地址

  https://console.bce.baidu.com/ai/#/ai/ocr/overview/index

1.2.2 根据步骤进行设置即可

在这里插入图片描述

2.复制粘贴代码，进行调试

2.1 将工具类中的API_KEY 和 SECRET_KEY 更换为你申请的对应key值即可

https://console.bce.baidu.com/ai/#/ai/ocr/app/list
在这里插入图片描述

2.2 开发文档的地址

https://cloud.baidu.com/doc/OCR/s/1k3h7y3db

2.3 线上调试的地址

https://console.bce.baidu.com/tools/?_=1668425998119#/api?product=AI&project=%E6%96%87%E5%AD%97%E8%AF%86%E5%88%AB&parent=%E9%80%9A%E7%94%A8%E5%9C%BA%E6%99%AFOCR&api=rest%2F2.0%2Focr%2Fv1%2Faccurate_basic&method=post

2.4 工具类（更换秘钥可直接使用）

package com.ruoyi.utils;

/**
 * @author  zhuenci
 * @ClassName imageToTextUtils
 * Description: 识别图片中的文字信息
 * @date  2024/3/14 10:19
 * @version 1.0
 *
 *  * 需要添加依赖
 *  * <!-- https://mvnrepository.com/artifact/com.squareup.okhttp3/okhttp -->
 *  * <dependency>
 *  *     <groupId>com.squareup.okhttp3</groupId>
 *  *     <artifactId>okhttp</artifactId>
 *  *     <version>4.12.0</version>
 *  * </dependency>
*/

import com.alibaba.fastjson.JSON;
import com.baomidou.mybatisplus.core.toolkit.CollectionUtils;
import com.ruoyi.common.utils.StringUtils;
import lombok.extern.slf4j.Slf4j;
import okhttp3.*;
import org.json.JSONException;
import org.json.JSONObject;

import java.io.IOException;
import java.net.URLEncoder;
import java.nio.file.Files;
import java.nio.file.Paths;
import java.util.Base64;
import java.util.List;
import java.util.Map;
@Slf4j
class imageToTextUtils {
    public static final String API_KEY = "ObKZGDN8***886qg2I";
    public static final String SECRET_KEY = "gMsL3LhQ***QT61i";

    public static final String AUTH_TOKEN_URL = "https://aip.baidubce.com/oauth/2.0/token";

    public static final String FILE_TO_TEXT_URL = "https://aip.baidubce.com/rest/2.0/ocr/v1/accurate_basic?access_token=";

    public static final String WORD_RESULT = "words_result";
    public static final String WORDS = "words";

    public static final String PDF_PAGE_NUM = "1";

    public static final String PDF_PAGE_SIZE = "pdf_file_size";




    static final OkHttpClient HTTP_CLIENT = new OkHttpClient().newBuilder().build();

    public static void main(String []args) {

        String text = getTextByPdfFile("D:\\**\\开发规范文档_V1.0.0.pdf", "1");
        String dealText = dealTextFormatString(text,true);
        String dealText2 = dealTextFormatString(text,false);
        List<Map> listByText = getListByText(text);
        Map<String, Object> textMap = getTextMap(text);
      /*  System.out.println("======================================================================");
        System.out.println(JSON.toJSONString(textMap));
        System.out.println("======================================================================");*/
         /*System.out.println(dealText2);
        System.out.println("======================================================================");*/

        String path = "D:\\**\\开发规范文档_V1.0.0.pdf";
        String multiPdfFileText = getMultiPdfFileText(path);
        System.out.println("=======================================================");
        System.out.println(multiPdfFileText);
        System.out.println("=======================================================");

    }

    /**
     * 返回的结果为追加值(标准化)
     * isHh(是否换行)
     **/
    public static String dealTextFormatString(String text,boolean isHh) {
        List<Map> list = getListByText(text);
        if(CollectionUtils.isEmpty(list)){
            return text;
        }
        StringBuilder sb = new StringBuilder();
        list.forEach(s->{
            if(!s.containsKey(WORDS)){
                return;
            }
            sb.append(s.get(WORDS));
            if(isHh){
                sb.append("\n");
            }
        });
        return sb.toString();
    }
    /**
     * 返回的结果为list<map>值
     **/
    public static List<Map> getListByText(String text){
        Map<String, Object> innerMap = getTextMap(text);
        if(CollectionUtils.isEmpty(innerMap) || !innerMap.containsKey(WORD_RESULT)){
            return null;
        }
        Object object = innerMap.get(WORD_RESULT);
        List<Map> list = JSON.parseArray(JSON.toJSONString(object), Map.class);
        return list;
    }

    /**
     * 返回的结果为Map<String,Object>
     **/
    public static Map<String, Object> getTextMap(String text){
        if(StringUtils.isEmpty(text)){
            return null;
        }
        com.alibaba.fastjson.JSONObject obj = JSON.parseObject(text);
        if(null == obj){
            return null;
        }
        Map<String, Object> innerMap = obj.getInnerMap();
        return innerMap;
    }

    /**
     * pdf 多页识别
     **/
    public static String getMultiPdfFileText(String path){
        String text = getTextByPdfFile(path, PDF_PAGE_NUM);
        if(StringUtils.isEmpty(text)){
            return text;
        }
        Map<String, Object> textMap = getTextMap(text);
        if(CollectionUtils.isEmpty(textMap) || !textMap.containsKey(WORD_RESULT)){
            return text;
        }
        Object object = textMap.get(PDF_PAGE_SIZE);
        int pageSize = Integer.parseInt(String.valueOf(object));
        StringBuilder sb = new StringBuilder();
        //此次获取的值进行格式化
        sb.append(dealTextFormatString(text,true));
        System.out.println("=============================第【1】页结束==========================");
        int next = Integer.parseInt(PDF_PAGE_NUM) + 1;
        return dealMultiText(sb,pageSize,path,String.valueOf(next));
    }

    private static String dealMultiText(StringBuilder sb, int pageSize, String path, String current) {
        String text = getTextByPdfFile(path, current);
        if(StringUtils.isEmpty(text)){
            return sb.toString();
        }
        Map<String, Object> textMap = getTextMap(text);
        if(CollectionUtils.isEmpty(textMap) || !textMap.containsKey(WORD_RESULT)){
            return  sb.toString();
        }
        //此次获取的值进行格式化
        sb.append(dealTextFormatString(text,true));
        int i = Integer.parseInt(current);
        System.out.println("=============================第【"+i+"】页结束==========================");
        if(pageSize > i){
            i = i + 1;
            try {
                Thread.sleep(500);
            }catch (Exception e){
                e.printStackTrace();
            }
            dealMultiText(sb,pageSize,path,String.valueOf(i));
        }
        return sb.toString();
    }


    /**
     * // pdf_file 可以通过 getFileContentAsBase64("C:\fakepath\项链发票.pdf") 方法获取,
     * 如果Content-Type是application/x-www-form-urlencoded时,二个参数传true
     *
     * 单页识别
     * @param pdfFilePath
     * @param pdfPageNum 识别的第几页
     * @return
     */
    public static String getTextByPdfFile(String pdfFilePath,String pdfPageNum){
        String result = StringUtils.EMPTY;
        try {
            if(StringUtils.isEmpty(pdfPageNum)){
                pdfPageNum = PDF_PAGE_NUM;
            }
            MediaType mediaType = MediaType.parse("application/x-www-form-urlencoded");
            String pdfFileData = getFileContentAsBase64(pdfFilePath, true);
            String param = String.format("pdf_file=%s&pdf_file_num=%s&detect_direction=false&paragraph=false&probability=false",pdfFileData,pdfPageNum);
            RequestBody body = RequestBody.create(mediaType, param);
            Request request = new Request.Builder()
                .url(FILE_TO_TEXT_URL + getAccessToken())
                .method("POST", body)
                .addHeader("Content-Type", "application/x-www-form-urlencoded")
                .addHeader("Accept", "application/json")
                .build();
            Response response = HTTP_CLIENT.newCall(request).execute();
            result = response.body().string();
        }catch (IOException e){
            log.error("获取pdf返回文字信息出错，错误信息为：{}",e.getMessage());
            e.printStackTrace();
        }
        log.info("返回结果为："+ result );
        return result;
    }

    /**
     * image 可以通过 getFileContentAsBase64("C:\fakepath\621eec7950394a008cc134ce8c797108.png") 方法获取,
     * 如果Content-Type是application/x-www-form-urlencoded时,第二个参数传true
     * @param imageUrl
     * @return
     */
    public static String getTextByImage(String imageUrl){
        String result = StringUtils.EMPTY;
        try {
            String imageData = getFileContentAsBase64(imageUrl, true);
            MediaType mediaType = MediaType.parse("application/x-www-form-urlencoded");
            String imageParam = String.format("image=%s&detect_direction=false&paragraph=false&probability=false",imageData);
            RequestBody body = RequestBody.create(mediaType, imageParam);
            Request request = new Request.Builder()
                .url(FILE_TO_TEXT_URL + getAccessToken())
                .method("POST", body)
                .addHeader("Content-Type", "application/x-www-form-urlencoded")
                .addHeader("Accept", "application/json")
                .build();
            Response response = HTTP_CLIENT.newCall(request).execute();
            result = response.body().string();

        }catch (IOException e){
            log.error("获取图片返回文字信息出错，错误信息为：{}",e.getMessage());
            e.printStackTrace();
        }
        log.info("返回结果为："+ result );
        return result;
    }

    /**
     * 获取文件base64编码
     *
     * @param path      文件路径
     * @param urlEncode 如果Content-Type是application/x-www-form-urlencoded时,传true
     * @return base64编码信息，不带文件头
     */
    static String getFileContentAsBase64(String path, boolean urlEncode) {
        String base64 = StringUtils.EMPTY;
        try {
            byte[] b = Files.readAllBytes(Paths.get(path));
            base64 = Base64.getEncoder().encodeToString(b);
            if (urlEncode) {
                base64 = URLEncoder.encode(base64, "utf-8");
            }
        }catch (IOException e){
            log.error("获取文件base64编码出现问题，异常信息为：{}",e.getMessage());
            e.printStackTrace();
        }
        return base64;
    }


    /**
     * 从用户的AK，SK生成鉴权签名（Access Token）
     *
     * @return 鉴权签名（Access Token）
     */
    static String getAccessToken() {
        String accessToken = StringUtils.EMPTY;
        try {
            MediaType mediaType = MediaType.parse("application/x-www-form-urlencoded");
            RequestBody body = RequestBody.create(mediaType, "grant_type=client_credentials&client_id=" + API_KEY + "&client_secret=" + SECRET_KEY);
            Request request = new Request.Builder()
                .url(AUTH_TOKEN_URL)
                .method("POST", body)
                .addHeader("Content-Type", "application/x-www-form-urlencoded")
                .build();
            Response response = HTTP_CLIENT.newCall(request).execute();
            accessToken = new JSONObject(response.body().string()).getString("access_token");
        }catch (IOException e){
            log.error("http请求出错，异常信息为：{}",e.getMessage());
            e.printStackTrace();
        }catch (JSONException e){
            log.error("获取accessToken出现错误，异常信息为：{}",e.getMessage());
            e.printStackTrace();
        }
        return accessToken;
    }

}