【百度OCR 封裝篇】OCR封裝隻IOCR自定義模版或分類器封裝兩種調用方式

全部代碼下載下傳位址：https://download.csdn.net/download/qq_33333654/11914449
package com.jeeplus.ocr;

import com.alibaba.fastjson.JSON;
import com.baidu.aip.ocr.AipOcr;
import com.baidu.aip.util.Base64Util;
import com.jeeplus.common.utils.CacheUtils;
import com.jeeplus.common.utils.DateUtils;
import com.jeeplus.common.utils.HttpUtil;
import com.jeeplus.common.utils.StringUtils;
import com.jeeplus.modules.sys.common.Constants;
import com.jeeplus.ocr.bean.AccessTokenResultBean;
import com.jeeplus.ocr.bean.AipOcrBasicGeneralResultBean;
import com.jeeplus.ocr.bean.AipOcrCustomResultBean;
import org.json.JSONObject;

import java.io.*;
import java.net.HttpURLConnection;
import java.net.URL;
import java.net.URLEncoder;
import java.util.*;

/**
 * @ProjectName: Hematopathy
 * @Package: com.jeeplus.ocr
 * @ClassName: AipOcrUtil
 * @Author: MC
 * @Description: ${description}
 * @Date: 2019/10/23 0023 10:48
 * @Version: 1.0
 */
public class AipOcrUtil {

    /**
     * @Method 通用文字識别
     * @Author MC
     path:本地檔案路徑
    options:請求API參數，可為空
     * @Return
     * @Date 2019/10/23 0023 11:36
     */
    public static AipOcrBasicGeneralResultBean basicGeneralByPath(String path, HashMap<String, String> options){
        if(checkFileIsImg(path)){
            if(options == null || options.isEmpty()){
                options = new HashMap<String, String>();
            }
            options.put("language_type", "CHN_ENG");
            options.put("detect_direction", "true");
            options.put("detect_language", "true");
            options.put("probability", "true");
            AipOcr client = AipOcrSingleton.getAipOcr();
            JSONObject res = client.basicGeneral(path, options);
            String resStr = res.toString();
            AipOcrBasicGeneralResultBean bean = JSON.parseObject(resStr, AipOcrBasicGeneralResultBean.class);
            return bean;
        }
        return null;
    }

    /**
     * @Method 通用文字識别
     * @Author MC
    url:網絡圖檔位址
    options:請求API參數，可為空
     * @Return
     * @Date 2019/10/23 0023 11:36
     */
    public static AipOcrBasicGeneralResultBean basicGeneralByURL(String url, HashMap<String, String> options){
        if(!StringUtils.isEmpty(url)){
            if(options == null || options.isEmpty()){
                options = new HashMap<String, String>();
            }
            options.put("language_type", "CHN_ENG");
            options.put("detect_direction", "true");
            options.put("detect_language", "true");
            options.put("probability", "true");
            AipOcr client = AipOcrSingleton.getAipOcr();
            JSONObject res = client.basicGeneralUrl(url, options);
            String resStr = res.toString();
            AipOcrBasicGeneralResultBean bean = JSON.parseObject(resStr, AipOcrBasicGeneralResultBean.class);
            return bean;
        }
        return null;
    }


    /**
     * @Method 根據模版ID或者分類器ID識别本地圖檔
     * @Author MC
    path：本地圖檔路徑
    templateSign：指定模版ID
    classifierId：指定分類器ID
    模版ID與分類ID至少一個不為空
     * @Return
     * @Date 2019/10/23 0023 13:59
     */
    public static AipOcrCustomResultBean customByPath(String path,String templateSign,String classifierId){
        if(checkFileIsImg(path)){
            if(StringUtils.isEmpty(templateSign) && StringUtils.isEmpty(classifierId)){
                return null;
            }
            HashMap<String, String> options = new HashMap<String, String>();
            if(!StringUtils.isEmpty(templateSign)){
                options.put("templateSign", templateSign);
            }
            if(!StringUtils.isEmpty(classifierId)){
                options.put("classifierId", classifierId);
            }
            AipOcr client = AipOcrSingleton.getAipOcr();
            JSONObject res = client.custom(path, options);
            String resStr = res.toString();
            AipOcrCustomResultBean bean = JSON.parseObject(resStr, AipOcrCustomResultBean.class);
            return bean;
        }
        return null;
    }

    /**
     * @Method 使用API調用方式進行自定義模版或者分類器的文字識别
     * @Author MC
     ak:app_key 必傳
    sk:secret_key 必傳
    filePath:本地檔案路徑 要求為圖檔格式 必傳
    templateSign：自定義模版ID
    classifierParams：自定義分類器ID
    templateSign和classifierParams至少傳遞一個
     * @Return
     * @Date 2019/10/24 0024 11:34
     */
    public static AipOcrCustomResultBean customByApiUrl(String ak, String sk,String filePath,String templateSign,String classifierParams) throws Exception {
        if(StringUtils.isEmpty(ak) || StringUtils.isEmpty(sk) || StringUtils.isEmpty(filePath)){
            return null;
        }
        if(StringUtils.isEmpty(templateSign) && StringUtils.isEmpty(classifierParams)){
            return null;
        }
        if(checkFileIsImg(filePath)){
            String imgStr = imgFile2Bytes2Encode(filePath);
            String param = "";
            if(!StringUtils.isEmpty(classifierParams)){
                param = "classifierId="+classifierParams+"&image=" + URLEncoder.encode(imgStr, "UTF-8");
            }else{
                param = "templateSign="+templateSign+"&image=" + URLEncoder.encode(imgStr, "UTF-8");
            }
            AccessTokenResultBean auth = getAuth(ak, sk);
            if(auth == null || !StringUtils.isEmpty(auth.getError())){
                return null;
            }

            String accessToken = auth.getAccess_token();
            // 請求模闆識别
            String result = HttpUtil.ocrPost(Constants.OCR_API_URL_GET_RECOGNISE, accessToken, param);
            AipOcrCustomResultBean bean = JSON.parseObject(result, AipOcrCustomResultBean.class);
            return bean;
        }
        return null;
    }

    /**
     * 擷取API通路token
     * 該token有一定的有效期，需要自行管理，當失效時需重新擷取.
     * @param ak - 百度雲官網擷取的 API Key
     * @param sk - 百度雲官網擷取的 Securet Key
     * @return assess_token 示例：
     * "24.460da4889caad24cccdb1fea17221975.2592000.1491995545.282335-1234567"
     */
    public static AccessTokenResultBean getAuth(String ak, String sk) {
        AccessTokenResultBean bean = (AccessTokenResultBean) CacheUtils.get(Constants.OCR_CACHE_NAME,Constants.OCR_CACHE_ACCESS_TOKEN);
        if(bean != null){
            //檢查緩存中的token是否過期
            long expires_in = bean.getMyTimeOut().getTime();
            Date nowDate = new Date();
            long nowDateTime = nowDate.getTime();
            if(expires_in > nowDateTime){
                return bean;
            }else{
                CacheUtils.remove(Constants.OCR_CACHE_NAME,Constants.OCR_CACHE_ACCESS_TOKEN);
            }
        }

        // 擷取token位址
        String authHost = Constants.OCR_API_URL_GET_ACCESS_TOKEN + "?";
        String getAccessTokenUrl = authHost
                // 1. grant_type為固定參數
                + "grant_type=client_credentials"
                // 2. 官網擷取的 API Key
                + "&client_id=" + ak
                // 3. 官網擷取的 Secret Key
                + "&client_secret=" + sk;
        try {
            URL realUrl = new URL(getAccessTokenUrl);
            // 打開和URL之間的連接配接
            HttpURLConnection connection = (HttpURLConnection) realUrl.openConnection();
            connection.setRequestMethod("GET");
            connection.connect();
            // 擷取所有響應頭字段
            Map<String, List<String>> map = connection.getHeaderFields();
            // 周遊所有的響應頭字段
//            for (String key : map.keySet()) {
//                System.err.println(key + "--->" + map.get(key));
//            }
            // 定義 BufferedReader輸入流來讀取URL的響應
            BufferedReader in = new BufferedReader(new InputStreamReader(connection.getInputStream()));
            String result = "";
            String line;
            while ((line = in.readLine()) != null) {
                result += line;
            }
            /**
             * 傳回結果示例
             */
            System.err.println("result:" + result);
            AccessTokenResultBean accessTokenResultBean = JSON.parseObject(result, AccessTokenResultBean.class);
            if(StringUtils.isEmpty(accessTokenResultBean.getError())){//擷取成功 進行緩存
                Integer expires_in = accessTokenResultBean.getExpires_in();
                if(expires_in > 0){
                    int days = expires_in / ( 60 * 60 * 24);
                    Date date=new Date();
                    Calendar calendar = new GregorianCalendar();
                    calendar.setTime(date);
                    calendar.add(Calendar.DATE,days);//把日期往後增加一天.整數往後推,負數往前移動
                    date=calendar.getTime();
                    accessTokenResultBean.setMyTimeOut(date);//設定有效期
                }else{
                    CacheUtils.remove(Constants.OCR_CACHE_NAME,Constants.OCR_CACHE_ACCESS_TOKEN);
                    return null;
                }
                CacheUtils.put(Constants.OCR_CACHE_NAME,Constants.OCR_CACHE_ACCESS_TOKEN,accessTokenResultBean);
            }else{
                CacheUtils.remove(Constants.OCR_CACHE_NAME,Constants.OCR_CACHE_ACCESS_TOKEN);
            }
            return accessTokenResultBean;
           /* JSONObject jsonObject = new JSONObject(result);
            String access_token = jsonObject.getString("access_token");
            return access_token;*/
        } catch (Exception e) {
            System.err.printf("擷取token失敗！");
            CacheUtils.remove(Constants.OCR_CACHE_NAME,Constants.OCR_CACHE_ACCESS_TOKEN);
            e.printStackTrace(System.err);
        }
        return null;
    }

    private static boolean checkFileIsImg(String path){
        if(StringUtils.isEmpty(path)){
            return false;
        }
        File srcFile = new File(path);
        // 判斷源檔案是否存在
        if (!srcFile.exists()) {
            return false;
        }
        if(path.indexOf(".") == -1){
            return false;
        }
        String suffix= path.substring(path.indexOf("."), path.length());
        if(
                suffix.equalsIgnoreCase("jpg") ||
                        suffix.equalsIgnoreCase("PNG ") ||
                        suffix.equalsIgnoreCase("GIF") ||
                        suffix.equalsIgnoreCase("tif") ||
                        suffix.equalsIgnoreCase("bmp")
        ){
            return false;
        }

        return true;
    }


    /**
     * @Method 圖檔檔案轉換為位元組在進行base64轉碼
     * @Author MC

     * @Return
     * @Date 2019/10/24 0024 10:19
     */
    public static String imgFile2Bytes2Encode(String filePath) throws IOException {
        if(!checkFileIsImg(filePath)){
            return null;
        }
        byte[] imgData = readFileByBytes(filePath);
        String imgStr = Base64Util.encode(imgData);
        return imgStr;
    }

    /**
     * @Method 檔案轉換為位元組在進行base64轉碼
     * @Author MC

     * @Return
     * @Date 2019/10/24 0024 10:19
     */
    public static String file2Bytes2Encode(String filePath) throws IOException {
        if(StringUtils.isEmpty(filePath)){
            return null;
        }
        File srcFile = new File(filePath);
        // 判斷源檔案是否存在
        if (!srcFile.exists()) {
            return null;
        }
        byte[] imgData = readFileByBytes(filePath);
        String imgStr = Base64Util.encode(imgData);
        return imgStr;
    }

    /**
     * 根據檔案路徑讀取byte[] 數組
     */
    public static byte[] readFileByBytes(String filePath) throws IOException {
        File file = new File(filePath);
        if (!file.exists()) {
            throw new FileNotFoundException(filePath);
        } else {
            ByteArrayOutputStream bos = new ByteArrayOutputStream((int) file.length());
            BufferedInputStream in = null;

            try {
                in = new BufferedInputStream(new FileInputStream(file));
                short bufSize = 1024;
                byte[] buffer = new byte[bufSize];
                int len1;
                while (-1 != (len1 = in.read(buffer, 0, bufSize))) {
                    bos.write(buffer, 0, len1);
                }

                byte[] var7 = bos.toByteArray();
                return var7;
            } finally {
                try {
                    if (in != null) {
                        in.close();
                    }
                } catch (IOException var14) {
                    var14.printStackTrace();
                }

                bos.close();
            }
        }
    }




}
【百度OCR 封裝篇】OCR封裝隻IOCR自定義模版或分類器封裝兩種調用方式

繼續閱讀

五十八、UiPath如何調用百度OCR（轉）

Python驗證碼識别tesseract-ocr安裝，報錯解決

tess4j的簡單使用doOCR異常彙總：

java實作阿裡雲圖檔文字識别

圖檔文字識别工具調研2.tesseract安裝與使用3.python

LEADTOOLS WinRT OCR識别代碼示例

場景文本檢測（一）-可微分二值化在基于語義分割方法的場景文本檢測中的應用動機和Contributions現存方法Methodology總結Ref

文本檢測算法----DB、DBNet1. 摘要2. 算法3. 實驗4. 結論

使用MODI（Microsoft Office Document Imaging）識别中文，但無法區分段落

主流深度學習OCR文字識别方法對比：Tesseract（LSTM）、CTPN+CRNN、Densenetopencv mser算法框出圖檔文字區域

【基于WPF+OneNote+Oracle的中文圖檔識别系統階段總結】之篇三：批量處理後的txt檔案入庫處理篇三：批量處理後的txt檔案入庫處理

百度雲 OCR 識别圖檔驗證碼

爬蟲驗證碼識别_工具篇：安裝pytesseract&Tesseract-OCR

應用移動端銀行卡識别技術，實作APP端快捷綁定銀行卡号

手寫字元識别

Jaspersfot Studio Create Check OCR Font
