Java实现识别发票信息

Java实现调用第三方接口识别发票信息

需求：对每个发票图片文件进行重命名，名称为发票号+固定信息，主要处理增值税发票

这里需要用到第三方接口，OCR识别功能，这里我用的是百度云接口，所以你需要注册百度云账号，然后生成密钥，获取权限接口！

百度云官方地址：https://console.bce.baidu.com/

百度云服务注册步骤

• 登陆百度云账号

  扫码或手机号码短信登陆，搜索文字识别服务
  

• 创建应用，生成密钥

  我这里只勾选一个增值税发票识别
  

• 获取密钥

  创建完应用后直接用就行了
  

代码实现步骤

• 通过client_id和client_secret获取token
• 通过token和相关参数请求识别发票接口
• 返回结果
• 处理文件

package com.shac.sellinvoiceautomation;

import com.alibaba.fastjson.JSON;
import com.alibaba.fastjson.JSONObject;
import lombok.extern.slf4j.Slf4j;
import okhttp3.*;
import org.junit.Test;
import org.junit.runner.RunWith;
import org.springframework.boot.test.context.SpringBootTest;
import org.springframework.test.context.junit4.SpringRunner;
import org.springframework.test.context.web.WebAppConfiguration;
import org.springframework.util.ObjectUtils;

import java.io.*;
import java.net.URLEncoder;
import java.nio.file.Files;
import java.nio.file.Paths;
import java.util.Base64;
import java.util.Objects;
import java.util.concurrent.TimeUnit;

/**
 * @ClassName OcrIdentificationInvoiceTest
 * @Description TODO
 * @Author jeckwu
 * @Date 2023/5/19 14:22
 */

@SpringBootTest
@RunWith(SpringRunner.class)
@WebAppConfiguration
@Slf4j
public class OcrIdentificationInvoiceTest {

    static final OkHttpClient HTTP_CLIENT = new OkHttpClient().newBuilder()
            .connectTimeout(15, TimeUnit.SECONDS)
            .readTimeout(15, TimeUnit.SECONDS)
            .writeTimeout(15, TimeUnit.SECONDS).build();

    // 获取token
    public String getToken() throws IOException {
        MediaType mediaType = MediaType.parse("application/json");
        RequestBody body = RequestBody.create(mediaType, "");
        Request request = new Request.Builder()
                .url("https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id=你的client_id&client_secret=你的client_secret")
                .method("POST", body)
                .addHeader("Content-Type", "application/json")
                .addHeader("Accept", "application/json")
                .build();
        Response response = HTTP_CLIENT.newCall(request).execute();
        assert response.body() != null;
        JSONObject jsonObject = JSON.parseObject(response.body().string());
        return jsonObject.getString("access_token");
    }

    // 识别发票
    public String ocrInvoice(String token,String imgBase64) throws IOException{
        MediaType mediaType = MediaType.parse("application/x-www-form-urlencoded");
        //如果Content-Type是application/x-www-form-urlencoded时,第二个参数传true
        RequestBody body = RequestBody.create(mediaType, "image="+imgBase64);
        Request request = new Request.Builder()
                .url("https://aip.baidubce.com/rest/2.0/ocr/v1/vat_invoice?access_token="+token)
                .method("POST", body)
                .addHeader("Content-Type", "application/x-www-form-urlencoded")
                .addHeader("Accept", "application/json")
                .build();
        Response response = HTTP_CLIENT.newCall(request).execute();
        assert response.body() != null;
        JSONObject jsonObject = JSON.parseObject(response.body().string());
        if (!ObjectUtils.isEmpty(jsonObject.getString("error_msg"))){
            log.error("识别失败，不是电子发票!msg:{}",jsonObject.getString("error_msg"));
            return null;
        }
        return jsonObject.getJSONObject("words_result").getString("InvoiceNum");
    }

    /**
     * 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
     * @throws IOException IO异常
     */
    static String getFileContentAsBase64(String path, boolean urlEncode) throws IOException {
        byte[] b = Files.readAllBytes(Paths.get(path));
        String base64 = Base64.getEncoder().encodeToString(b);
        if (urlEncode) {
            base64 = URLEncoder.encode(base64, "utf-8");
        }
        return base64;
    }
  
    final String invoicePath = "/Users/jeckwu/Desktop/test/IMG_8140.JPG";

    @Test
    public void test01() throws IOException {
        String token = getToken();
        String invoiceNumber = ocrInvoice(token, getFileContentAsBase64(invoicePath, true));
        if (!ObjectUtils.isEmpty(invoiceNumber))
            log.info("照片路径:{},发票号码:{}",invoicePath,invoiceNumber);
        else {
            log.info("照片路径:{},识别失败！不是电子发票！",invoicePath);
        }
    }

    final String OCR_FOLDER = "/Users/jeckwu/Desktop/test/";
    final String FILE_NAME_PREFIX = "test-invoice-";

    // 批量识别
    @Test
    public void test02() throws IOException {
        File file = new File(OCR_FOLDER);
        if (ObjectUtils.isEmpty(file.listFiles())) {
            log.info("没有要识别的发票文件！！！");
            return;
        }
        String token = getToken();
        // 遍历根目录的文件夹
        for (File listFile : Objects.requireNonNull(file.listFiles())) {
            // 是否是文件
            if (listFile.isFile()) {
                if (listFile.getAbsolutePath().indexOf(FILE_NAME_PREFIX)>0) continue;
                // 识别发票文件
                String invoiceNum= ocrInvoice(token, getFileContentAsBase64(listFile.getAbsolutePath(), true));
                if (ObjectUtils.isEmpty(invoiceNum)) {
                    log.info("照片路径:{},识别失败！不是电子发票！",listFile.getAbsolutePath());
                    continue;
                }
                // 修改文件名
                listFile.renameTo(new File(listFile.getParent() +File.separator+ FILE_NAME_PREFIX + invoiceNum + ".jpg"));
                log.info("发票路径:{},发票号码:{},识别成功",listFile.getAbsolutePath(),invoiceNum);
            }else if (listFile.isDirectory()){
                // 遍历有发票文件的文件夹
                File[] files = listFile.listFiles();
                if (ObjectUtils.isEmpty(files)) continue;
                for (File invoiceFile : files) {
                    if (invoiceFile.getAbsolutePath().indexOf(FILE_NAME_PREFIX)>0) continue;
                    // 识别发票文件
                    String invoiceNum= ocrInvoice(token, getFileContentAsBase64(invoiceFile.getAbsolutePath(), true));
                    if (ObjectUtils.isEmpty(invoiceNum)) {
                        log.info("照片路径:{},识别失败！不是电子发票！",invoiceFile.getAbsolutePath());
                        continue;
                    }
                    File tempFile = new File(invoiceFile.getParent() +File.separator+ FILE_NAME_PREFIX + invoiceNum + ".jpg");
                    // 修改文件名
//                    copyFileUsingStream(invoiceFile,tempFile);
                    invoiceFile.renameTo(tempFile);
                    log.info("发票路径:{},发票号码:{},识别成功",invoiceFile.getAbsolutePath(),invoiceNum);
                }
            }
        }
    }

    // 复制文件
    private static void copyFileUsingStream(File source, File dest) throws IOException {
        InputStream is = null;
        OutputStream os = null;
        try {
            is = new FileInputStream(source);
            os = new FileOutputStream(dest);
            byte[] buffer = new byte[1024];
            int length;
            while ((length = is.read(buffer)) > 0) {
                os.write(buffer, 0, length);
            }
        } finally {
            assert is != null;
            is.close();
            assert os != null;
            os.close();
        }
    }

}

其中百度云这个识别服务每个月有免费试用次数且有限！！有需要可以去了解详情。

识别效果

基本能识别出来，前提是发票必须清楚，识别信息也比较全面，具体可参考该接口文档。

百度云官方调试网站：https://console.bce.baidu.com/tools/#/api?product=AI&project=%E6%96%87%E5%AD%97%E8%AF%86%E5%88%AB&parent=%E8%B4%A2%E5%8A%A1%E7%A5%A8%E6%8D%AEOCR&api=rest%2F2.0%2Focr%2Fv1%2Fvat_invoice&method=post

如果文章对你有帮助请点赞支持！！