后端获取数据：

在ScoresController中的addd：

这个方法主要完成的工作包括：

1、解析前端传来的json，包括该课程的课程编号cs_id、课序号les_ord、图片对象数组imgimg。然后再对图片数组进行处理，提取数组中的图片对象对应的base64信息。

因为前端传的数据imgimg也是字符串格式，所以用字符串切割的处理方式，用picbase64=作为分隔符来进行字符串切割，进而获取每个图片的base64；

2、将处理后的base64信息图片传递给识别接口SampleTwo中的identify方法，该方法再完成识别的同时进行成绩的计算工作，然后将成绩返回给addd方法；

3、再由addd方法完成数据库update的调用，进而完成学生成绩的插入工作。

    /***   添加(图片上传的addd)** @param map* @return*/@AutoLog(value = "临时的临时成绩-添加-图片上传")@ApiOperation(value="临时的临时成绩-添加-图片上传", notes="临时的临时成绩-添加-图片上传")@PostMapping(value = "/addd")    public Result<String> addd(@RequestBody Map<String,Object> map) {//scoresService.save(scores);System.out.println("-----------test图片上传--------");String url = new String();String cs_id = new String();String les_ord = new String();String ScrMeth = new String();List<Pic> base64arr = new ArrayList<Pic>();Pic pic = new Pic();String base64 = "";int picnum = 0;String substr[] = new String[10];String picbase64[] = new String[10];if (map.containsKey("url")){url = map.get("url").toString();System.out.println(url);}if (map.containsKey("filename")){String filename = map.get("filename").toString();System.out.println(filename);}if (map.containsKey("imgimg")){Object list = map.get("imgimg");String liststr = list.toString();//System.out.println(liststr);String str[] = liststr.split("picuid");picnum = str.length-1;System.out.println("picnum = "+picnum);for(int i=0;i<picnum;i++){substr[i] = liststr.split("picbase64=")[i+1];picbase64[i] = substr[i].split("}")[0];// System.out.println(picbase64[i]);}}if(map.containsKey("cs_id")){cs_id = map.get("cs_id").toString();}if(map.containsKey("les_ord")){les_ord = map.get("les_ord").toString();}if (map.containsKey("ScrMeth")){ScrMeth = map.get("ScrMeth").toString();System.out.println(ScrMeth);}System.out.println("获取成功");
//       return "success";System.out.println("------识别(不带位置）------");String path_pic = MinioUtil.path_pic;System.out.println("path_pic in system= "+path_pic);//成绩比例处理int usualScores=-1;int testScores=-1;int scoreNum=-1;if(this.temporaryScores.getTestPercent()==null){this.temporaryScores.setTestPercent(0);}if(this.temporaryScores.getUsualPercent()==null){this.temporaryScores.setUsualPercent(0);}double test_percent=this.temporaryScores.getTestPercent()*0.01;double usual_percent = this.temporaryScores.getUsualPercent()*0.01;for(int j=0;j<picnum;j++) {String pic64 = picbase64[j];SampleTwo.ChangeForm(pic64);String path65 = "D:/aaa.jpg";List<List<Object>> Slist = SampleTwo.identify(path65,test_percent,usual_percent);//List<List<Object>> Slist = SampleTwo.identify(path_pic);//识别结果处理List<Object> scorelist = new ArrayList<Object>();Scores scores = new Scores();int Sl_lenth = Slist.size();for (int i = 0; i < Sl_lenth; i++) {scorelist = Slist.get(i);String stu_name = (String) scorelist.get(0);String stu_id = (String) scorelist.get(1);int exam_score = (int) scorelist.get(2);int usual_score = (int) scorelist.get(3);int test_score = (int) scorelist.get(4);int scr = (int) scorelist.get(5);String score = scr + "";if (ScrMeth.equals("评级制")) {score = UtilClass.ScrChange(score);}boolean a = scoresService.updateExamScores(exam_score, cs_id, les_ord, stu_id);boolean b = scoresService.updateUsualScores(usual_score, cs_id, les_ord, stu_id);boolean c = scoresService.updateTestScores(test_score, cs_id, les_ord, stu_id);boolean d = scoresService.updateScores(score, cs_id, les_ord, stu_id);if (a && b && c && d) {System.out.println(stu_id + " " + stu_name + " 同学的成绩插入成功");}}}return Result.OK("添加成功！");}

    <select id="updateUsualScores"  resultType="org.jeecg.modules.demo.ScoresInput.entity.Scores">update scoressetusual_Scores = #{usual_Scores}wherecs_id = #{cs_id} and stu_id=#{stu_id} and les_ord=#{les_ord}</select><select id="updateExamScores" resultType="org.jeecg.modules.demo.ScoresInput.entity.Scores">update scoressetexam_Scores = #{exam_Scores}wherecs_id = #{cs_id} and stu_id=#{stu_id} and les_ord=#{les_ord}</select><select id="updateScores"  resultType="org.jeecg.modules.demo.ScoresInput.entity.Scores">update scoressetscores = #{scores}wherecs_id = #{cs_id} and stu_id=#{stu_id} and les_ord=#{les_ord}</select><select id="updateTestScores" resultType="org.jeecg.modules.demo.ScoresInput.entity.Scores">update scoressettest_scores = #{test_scores}wherecs_id = #{cs_id} and stu_id=#{stu_id} and les_ord=#{les_ord}</select>

从前端获取到的图片的base64信息需要进行格式转换，方法如下：

    public static String ChangeForm(String basestr){String data = basestr;// System.out.println("data = "+data);String base64Data =  data.split(",")[1];// System.out.println("base64Data = "+base64Data);/*** 2.解码成字节数组*/Base64.Decoder decoder = Base64.getDecoder();byte[] bytes = decoder.decode(base64Data);/*** 3.字节流转文件*/FileOutputStream fos = null;try {try {fos = new FileOutputStream("D:\\aaa.jpg");} catch (FileNotFoundException e) {e.printStackTrace();}try {fos.write(bytes);} catch (IOException e) {e.printStackTrace();}return base64Data;} finally {if (fos != null){try {fos.close();} catch (IOException e) {e.printStackTrace();}}}}

识别结果处理：

上传的图片经过识别后是json格式的字符串，所以下面可以根据字符串的处理方式来对获得结果进行处理，进而获取我们想要的信息；

SampleTwo中的identity方法：

public class SampleTwo {//设置APPID/AK/SKpublic static final String APP_ID = "26041291";public static final String API_KEY = "3zSCXcXcHcw7KEHlk5SLOxug";public static final String SECRET_KEY = "z2HRbmgH7LNxH7AII2EmdLheKoHsj7aj";
……
}

identify方法主要完成的工作包括：

1、调用百度文字识别API，对ScoresControll中获得的图片进行识别，获取到识别结果的json形式的字符串；

2、对字符串进行处理：

需要考虑上传图片的格式不同的问题

3、将获取到的学生信息和对应成绩存入Scores实体的list中

再此过程中需要完成对成绩的计算工作，利用出入该方法的参数（平时成绩占比、实验成绩占比）算出学生的最终成绩score ；

    public static List<List<Object>> identify(String path,double test_percent,double usual_percent){// 初始化一个AipOcrAipOcr client = new AipOcr(APP_ID, API_KEY, SECRET_KEY);// 可选：设置网络连接参数client.setConnectionTimeoutInMillis(2000);client.setSocketTimeoutInMillis(60000);// 可选：设置代理服务器地址, http和socket二选一，或者均不设置//client.setHttpProxy("proxy_host", proxy_port);  // 设置http代理//client.setSocketProxy("proxy_host", proxy_port);  // 设置socket代理// 可选：设置log4j日志输出格式，若不设置，则使用默认配置// 也可以直接通过jvm启动参数设置此环境变量System.setProperty("aip.log4j.conf", "path/to/your/log4j.properties");// 调用接口JSONObject res = client.basicGeneral(path, new HashMap<String, String>());String string = "";List<List<Object>> Slist = new ArrayList<List<Object>>();try {int length = res.length();System.out.println("length = "+length);int num = res.getInt("words_result_num");// int num = Integer.parseInt(res.getString("word_result_num"));int row = num/5;System.out.println("num = "+num+", row = "+row);String s = res.toString();String str = s.substring(16);String strm = "";String stri = "";int m = 0;String stu_name = "";String stu_id = "";String score_exam = "";String score_usual = "";String score_lab = "";int sc_exam = 0;int sc_usual = 0;int sc_lab = 0;int score = 0;double a = test_percent;double b = usual_percent;double c = 1-a-b;String substr1 = str.split(":")[1];System.out.println("substr1 = "+substr1);String title = "报表";boolean withtitle = substr1.contains(title);System.out.println("withtitle = "+withtitle);String substr3 = str.split(":")[3];String header = "姓名";System.out.println("substr3 = "+substr3);boolean withheaderandtitle = substr3.contains(header);boolean withonlyheader = substr1.contains(header);for(int i=0;i<row-1;i++){System.out.println("------学生"+i+1+"-------");//strm = str.split(":")[i+1];// stri = strm.split("\"")[1];//System.out.println("strm = "+strm+"   stri = "+stri);for(int j=0;j<5;j++){System.out.println("------第"+j+1+"项-------"+j);//  String strj1 = str.split(":")[5*i+j+1+5];//   System.out.println("j = "+j);//   System.out.println("strj1 = "+strj1);String strj1 = "";if(withonlyheader){strj1 = str.split(":")[5*i+j+1+5];System.out.println("11"+strj1);}if(withtitle){strj1 = str.split(":")[5*i+j+1+7];System.out.println("22"+strj1);}if(!withtitle){if(!withonlyheader){strj1 = str.split(":")[5*i+j+1];System.out.println("33"+strj1);}}String strj2 = strj1.split("\"")[1];//  System.out.println("strj1 = "+strj1+", strj2 = "+strj2);switch(j){case 0:{stu_name = strj2;//      System.out.println("stu_name = "+stu_name+", j = "+j);break;}case 1:{stu_id = strj2;//   System.out.println("stu_id = "+stu_id+", j = "+j);break;}case 2:{score_exam = strj2;//  System.out.println("score_exam = "+score_exam+", j = "+j);sc_exam = UtilClass.checkNum(strj2);// sc_exam = Integer.parseInt(score_exam);//  System.out.println("sc_exam = "+sc_exam);break;}case 3:{score_usual = strj2;// System.out.println("score_usual = "+score_usual+", j = "+j);sc_usual = UtilClass.checkNum(strj2);//sc_usual = Integer.parseInt(score_usual);// System.out.println("sc_usaul = "+sc_usual);break;}case 4:{score_lab = strj2;// System.out.println("score_lab = "+score_lab+", j = "+j);sc_lab = UtilClass.checkNum(strj2);///sc_lab = Integer.parseInt(score_lab);// System.out.println("sc_lab = "+sc_lab);break;}}}double sc = 0;sc = a*sc_exam + b*sc_usual + c*sc_lab;score = (int) Math.round(sc);
//                score = Integer.parseInt(String.valueOf(sc));System.out.println("stu_name = "+stu_name+", stu_id = "+stu_id+", sc_exam = "+sc_exam+", sc_usual = "+sc_usual+", sc_lab = "+sc_lab+", score = "+score);List<Object> scoresList = new ArrayList<Object>();scoresList.add(0,stu_name);scoresList.add(1,stu_id);scoresList.add(2,sc_exam);scoresList.add(3,sc_usual);scoresList.add(4,sc_lab);scoresList.add(5,score);Slist.add(scoresList);}// System.out.println(res.toString(2));string = res.toString(2);} catch (JSONException e) {e.printStackTrace();}return Slist;}

IEMS_8_图片识别_2相关推荐

Python实现图片识别加翻译，高薪必备技能
Python使用百度AI接口实现图片识别加翻译 python python诞生30周年 encoding:utf-8 import requests import base64 from PIL im ...
物联网技术周报第 141 期: 使用 Alexa Voice 和 Raspberry Pi 构建图片识别应用
新闻 \\ \\t <Ubuntu 16.04 LTS 已在英特尔 NUC 和物联网主板上获得认证>在全球拥有数百万用户的热门GNU/Linux发行版本--Ubuntu的背后公司Canon ...
iOS--OCR图片识别
应公司财务需求,要做一个收据识别功能.所以在网上搜索了下三方SDK,其中tesseract-ocr受到了大多数网友的推荐.我当然是前往https://github.com/gali8/Tesserac ...
Atitit java 二维码识别图片识别
Atitit java 二维码识别图片识别 1.1. 解码1 1.2. 首先,我们先说一下二维码一共有40个尺寸.官方叫版本Version.1 1.3. 二维码的样例:2 1.4. 定位图案2 1. ...
plt保存图片_人工智能Keras CNN卷积神经网络的图片识别模型训练
CNN卷积神经网络是人工智能的开端,CNN卷积神经网络让计算机能够认识图片,文字,甚至音频与视频.CNN卷积神经网络的基础知识,可以参考:CNN卷积神经网络 LetNet体系结构是卷积神经网络的&qu ...
[验证码识别技术]-初级的滑动式验证图片识别
初级的滑动式验证图片识别方案 1 abstract 验证码作为一种自然人的机器人的判别工具,被广泛的用于各种防止程序做自动化的场景中.传统的字符型验证安全性已经名存实亡的情况下,各种新型的验证码如雨后 ...
DL之CNN：利用卷积神经网络算法(2→2,基于Keras的API-Functional)利用MNIST(手写数字图片识别)数据集实现多分类预测
DL之CNN:利用卷积神经网络算法(2→2,基于Keras的API-Functional)利用MNIST(手写数字图片识别)数据集实现多分类预测目录输出结果设计思路核心代码输出结果下边两张 ...
DL之CNN：利用卷积神经网络算法(2→2,基于Keras的API-Sequential)利用MNIST(手写数字图片识别)数据集实现多分类预测
DL之CNN:利用卷积神经网络算法(2→2,基于Keras的API-Sequential)利用MNIST(手写数字图片识别)数据集实现多分类预测目录输出结果设计思路核心代码输出结果 1.10 ...
DL之DNN：利用DNN【784→50→100→10】算法对MNIST手写数字图片识别数据集进行预测、模型优化
DL之DNN:利用DNN[784→50→100→10]算法对MNIST手写数字图片识别数据集进行预测.模型优化导读目的是建立三层神经网络,进一步理解DNN内部的运作机制目录输出结果设计思路 ...

IEMS_8_图片识别_2

后端获取数据：

在ScoresController中的addd：

识别结果处理：

SampleTwo中的identity方法：

IEMS_8_图片识别_2相关推荐

最新文章

热门文章