利用百度api实现ocr识别发票

神秘彩虹海

已于 2024-01-04 15:04:27 修改

阅读量1.2k

点赞数 2

分类专栏： java学习文章标签： ocr java

于 2023-06-13 16:30:36 首次发布

本文链接：https://blog.csdn.net/weixin_44219955/article/details/131190813

版权

java学习专栏收录该内容

14 篇文章 1 订阅

订阅专栏

一、准备工作

1、百度智能云申请一个文字识别的api
在这里插入图片描述
里面每个月有1000次调用测试应该够用

2、一个springboot项目

二、主要代码

①、pom依赖引入

 <dependency>
            <groupId>com.baidu.aip</groupId>
            <artifactId>java-sdk</artifactId>
            <version>4.16.5</version>
            <exclusions>
                <exclusion>
                    <artifactId>guava</artifactId>
                    <groupId>com.google.guava</groupId>
                </exclusion>
                <exclusion>
                    <artifactId>slf4j-simple</artifactId>
                    <groupId>org.slf4j</groupId>
                </exclusion>
            </exclusions>
        </dependency>
        <dependency>
            <groupId>com.baidubce</groupId>
            <artifactId>api-explorer-sdk</artifactId>
            <version>1.0.3.1</version>
            <exclusions>
                <exclusion>
                    <artifactId>guava</artifactId>
                    <groupId>com.google.guava</groupId>
                </exclusion>
            </exclusions>
        </dependency>
        <dependency>
            <groupId>com.squareup.okhttp3</groupId>
            <artifactId>okhttp</artifactId>
            <version>3.10.0</version>
        </dependency>

②、控制层

    @ApiOperation("识别发票")
    @ResponseBody
    public ResponseEntity<InvoiceInfo> identifyInvoice(MultipartFile file) throws BizException {
        return new ResponseEntity<>(invoiceInfoService.identifyInvoice(file), HttpStatus.OK);
    }

③、实现层

    public InvoiceInfo identifyInvoice(MultipartFile file) {
        InvoiceInfo invoiceInfo = new InvoiceInfo();
        // 请求url
        String url = "https://aip.baidubce.com/rest/2.0/ocr/v1/vat_invoice";
        try {
            byte[] fileData = file.getBytes();
            String originalFilename = file.getOriginalFilename();
            boolean isPdf = originalFilename.endsWith("pdf") || originalFilename.endsWith("PDF");

            String fileStr = Base64Util.encode(fileData);
            String fileParam = URLEncoder.encode(fileStr, "UTF-8");
            String param = (isPdf ? "pdf_file" : "image") + "=" + fileParam;
            // 注意这里仅为了简化编码每一次请求都去获取access_token，线上环境access_token有过期时间， 客户端可自行缓存，过期后重新获取。
            String accessToken = AuthService.getAuth();
            String result = HttpUtil.post(url, accessToken, param);
            JSONObject jsonObject = JSON.parseObject(result);
            if (jsonObject.getInteger("error_code") != null) {
                String errorMsg = BaiduOcrConstant.ERROR_CODE_MAP.get(jsonObject.getInteger("error_code"));
                System.out.println("票据识别失败，错误信息：" + errorMsg);
            } else {
                JSONObject words_result = jsonObject.getJSONObject("words_result");
                invoiceInfo.setInvoiceType(words_result.getString("InvoiceTypeOrg"));
                invoiceInfo.setImportType(isPdf ? 0L : 1L);
                invoiceInfo.setCheckFlag(0L);
                invoiceInfo.setSupplierName(words_result.getString("SellerName"));
                invoiceInfo.setSupplierNumber(words_result.getString("SellerRegisterNum"));
                invoiceInfo.setSupplierDescription(words_result.getString("SellerAddress"));
                invoiceInfo.setBankNumber(words_result.getString("SellerBank"));
                invoiceInfo.setPurchaserName(words_result.getString("PurchaserName"));
                invoiceInfo.setPurchaserNumber(words_result.getString("PurchaserRegisterNum"));
                invoiceInfo.setBillingDate(words_result.getDate("InvoiceDate"));
                invoiceInfo.setInvoiceCode(words_result.getString("InvoiceCodeConfirm"));
                invoiceInfo.setInvoiceNumber(words_result.getString("InvoiceNum"));
                invoiceInfo.setCheckCode(words_result.getString("CheckCode"));
                invoiceInfo.setAmountTotal(words_result.getBigDecimal("AmountInFiguers"));
                invoiceInfo.setAmount(words_result.getBigDecimal("TotalAmount"));
                if("*".equals(words_result.getString("TotalTax"))){
                    invoiceInfo.setTaxAmount(new BigDecimal("0.00"));
                }else {
                    invoiceInfo.setTaxAmount(words_result.getBigDecimal("TotalTax"));
                }
                invoiceInfo.setRate(words_result.getBigDecimal("TotalAmount"));
                invoiceInfo.setCreateTime(new Date());
            }
            return invoiceInfo;
        } catch (Exception e) {
            e.printStackTrace();
        }
        return null;
    }
}

其次就是百度需要的utils包

最后直接调用就行
在这里插入图片描述