JAVA使用Tess4J进行ocr识别
Tess4J是对Tesseract OCR API.的Java JNA 封装。使java能够通过调用Tess4J的API来使用Tesseract OCR。支持的格式:TIFF,JPEG,GIF,PNG,BMP,JPEG,and PDF
Tesseract 的github地址:https://github.com/tesseract-ocr/tesseract
Tess4J的github地址:https://github.com/nguyenq/tess4j
Tess4J API 提供的功能:
1、直接识别支持的文件
2、识别图片流
3、识别图片的某块区域
4、将识别结果保存为 TEXT/ HOCR/ PDF/ UNLV/ BOX
5、通过设置取词的等级,提取识别出来的文字
6、获得每一个识别区域的具体坐标范围
7、调整倾斜的图片
8、裁剪图片
9、调整图片分辨率
10、从粘贴板获得图像
11、克隆一个图像(目的:创建一份一模一样的图片,与原图在操作修改上,不相 互影响)
12、图片转换为二进制、黑白图像、灰度图像
13、反转图片颜色
demo.java:
/**
* Test of doOCR method, of class Tesseract.
* 根据图片文件进行识别
* @throws Exception while processing image.
*/
@Test
public void testDoOCR_File() throws Exception {
logger.info("doOCR on a jpg image");
File imageFile = new File(this.testResourcesDataPath, "ocr.png");
//set language
instance.setDatapath(testResourcesLanguagePath);
instance.setLanguage("chi_sim");
String result = instance.doOCR(imageFile);
logger.info(result);
}
/**
* Test of doOCR method, of class Tesseract.
* 根据图片流进行识别
* @throws Exception while processing image.
*/
@Test
public void testDoOCR_BufferedImage() throws Exception {
logger.info("doOCR on a buffered image of a PNG");
File imageFile = new File(this.testResourcesDataPath, "ocr.png");
BufferedImage bi = ImageIO.read(imageFile);
//set language
instance.setDatapath(testResourcesLanguagePath);
instance.setLanguage("chi_sim");
String result = instance.doOCR(bi);
logger.info(result);
}
/**
* Test of getSegmentedRegions method, of class Tesseract.
* 得到每一个划分区域的具体坐标
* @throws java.lang.Exception
*/
@Test
public void testGetSegmentedRegions() throws Exception {
logger.info("getSegmentedRegions at given TessPageIteratorLevel");
File imageFile = new File(testResourcesDataPath, "ocr.png");
BufferedImage bi = ImageIO.read(imageFile);
int level = TessPageIteratorLevel.RIL_SYMBOL;
logger.info("PageIteratorLevel: " + Utils.getConstantName(level, TessPageIteratorLevel.class));
List<Rectangle> result = instance.getSegmentedRegions(bi, level);
for (int i = 0; i < result.size(); i++) {
Rectangle rect = result.get(i);
logger.info(String.format("Box[%d]: x=%d, y=%d, w=%d, h=%d", i, rect.x, rect.y, rect.width, rect.height));
}
assertTrue(result.size() > 0);
}
/**
* Test of doOCR method, of class Tesseract.
* 根据定义坐标范围进行识别
* @throws Exception while processing image.
*/
@Test
public void testDoOCR_File_Rectangle() throws Exception {
logger.info("doOCR on a BMP image with bounding rectangle");
File imageFile = new File(this.testResourcesDataPath, "ocr.png");
//设置语言库
instance.setDatapath(testResourcesLanguagePath);
instance.setLanguage("chi_sim");
//划定区域
// x,y是以左上角为原点,width和height是以xy为基础
Rectangle rect = new Rectangle(84, 21, 15, 13);
String result = instance.doOCR(imageFile, rect);
logger.info(result);
}
/**
* Test of createDocuments method, of class Tesseract.
* 存储结果
* @throws java.lang.Exception
*/
@Test
public void testCreateDocuments() throws Exception {
logger.info("createDocuments for png");
File imageFile = new File(this.testResourcesDataPath, "ocr.png");
String outputbase = "target/test-classes/docrenderer-2";
List<RenderedFormat> formats = new ArrayList<RenderedFormat>(Arrays.asList(RenderedFormat.HOCR, RenderedFormat.TEXT));
//设置语言库
instance.setDatapath(testResourcesLanguagePath);
instance.setLanguage("chi_sim");
instance.createDocuments(new String[]{imageFile.getPath()}, new String[]{outputbase}, formats);
}
/**
* Test of getWords method, of class Tesseract.
* 取词方法
* @throws java.lang.Exception
*/
@Test
public void testGetWords() throws Exception {
logger.info("getWords");
File imageFile = new File(this.testResourcesDataPath, "ocr.png");
//设置语言库
instance.setDatapath(testResourcesLanguagePath);
instance.setLanguage("chi_sim");
//按照每个字取词
int pageIteratorLevel = TessPageIteratorLevel.RIL_SYMBOL;
logger.info("PageIteratorLevel: " + Utils.getConstantName(pageIteratorLevel, TessPageIteratorLevel.class));
BufferedImage bi = ImageIO.read(imageFile);
List<Word> result = instance.getWords(bi, pageIteratorLevel);
//print the complete result
for (Word word : result) {
logger.info(word.toString());
}
}
/**
* Test of Invalid memory access.
* 处理倾斜
* @throws Exception while processing image.
*/
@Test
public void testDoOCR_SkewedImage() throws Exception {
//设置语言库
instance.setDatapath(testResourcesLanguagePath);
instance.setLanguage("chi_sim");
logger.info("doOCR on a skewed PNG image");
File imageFile = new File(this.testResourcesDataPath, "ocr_skewed.jpg");
BufferedImage bi = ImageIO.read(imageFile);
ImageDeskew id = new ImageDeskew(bi);
double imageSkewAngle = id.getSkewAngle(); // determine skew angle
if ((imageSkewAngle > MINIMUM_DESKEW_THRESHOLD || imageSkewAngle < -(MINIMUM_DESKEW_THRESHOLD))) {
bi = ImageHelper.rotateImage(bi, -imageSkewAngle); // deskew image
}
String result = instance.doOCR(bi);
logger.info(result);
}
Tess4JDemo码云地址:https://gitee.com/zhaohuihbwj/Tess4JDemo
Java OCR文字识别(Tess4J)
OCR (Optical Character Recognition,光学字符识别)是指电子设备(例如扫描仪或数码相机)检查纸上打印的字符,通过检测暗、亮的模式确定其形状,然后用字符识别方法将形状翻译成计算机文字的过程;即,针对印刷体字符,采用光学的方式将纸质文档中的文字转换成为黑白点阵的图像文件,并通过识别软件将图像中的文字转换成文本格式,供文字处理软件进一步编辑加工的技术。如何除错或利用辅助信息提高识别正确率,是OCR最重要的课题,ICR(Intelligent Character Recognition)的名词也因此而产生。衡量一个OCR系统性能好坏的主要指标有:拒识率、误识率、识别速度、用户界面的友好性,产品的稳定性,易用性及可行性等。
Tess4J是对google tesseract ocr的java库的一种实现
1.maven添加依赖
<!-- https://mvnrepository.com/artifact/net.sourceforge.tess4j/tess4j -->
<dependency>
<groupId>net.sourceforge.tess4j</groupId>
<artifactId>tess4j</artifactId>
<version>3.2.1</version>
</dependency>
2.工具类编辑
/**
* tesseract for java, ocr(Optical Character Recognition,光学字符识别)
* 工具类
* @author wind
*/
public class Tess4jUtils {
/**
* 从图片中提取文字,默认设置英文字库,使用classpath目录下的训练库
* @param path
* @return
*/
public static String readChar(String path){
// JNA Interface Mapping
ITesseract instance = new Tesseract();
// JNA Direct Mapping
// ITesseract instance = new Tesseract1();
File imageFile = new File(path);
//In case you don't have your own tessdata, let it also be extracted for you
//这样就能使用classpath目录下的训练库了
File tessDataFolder = LoadLibs.extractTessResources("tessdata");
//Set the tessdata path
instance.setDatapath(tessDataFolder.getAbsolutePath());
//英文库识别数字比较准确
instance.setLanguage(Const.ENG);
return getOCRText(instance, imageFile);
} /**
* 从图片中提取文字
* @param path 图片路径
* @param dataPath 训练库路径
* @param language 语言字库
* @return
*/
public static String readChar(String path, String dataPath, String language){
File imageFile = new File(path);
ITesseract instance = new Tesseract();
instance.setDatapath(dataPath);
//英文库识别数字比较准确
instance.setLanguage(language);
return getOCRText(instance, imageFile);
} /**
* 识别图片文件中的文字
* @param instance
* @param imageFile
* @return
*/
private static String getOCRText(ITesseract instance, File imageFile){
String result = null;
try {
result = instance.doOCR(imageFile);
} catch (TesseractException e) {
e.printStackTrace();
}
return result;
} public static void main(String[] args) {
/*String path = "src/main/resources/image/text.png";
System.out.println(readChar(path));*/ String ch = "src/main/resources/image/ch.png";
System.out.println(readChar(ch, "src/main/resources", Const.CHI_SIM));
} }
注释:中文结果不够准确,需要自己训练字库
具体训练字库,以及完整代码请移步https://github.com/followwwind/javautils
JAVA使用Tess4J进行ocr识别的更多相关文章
- java 使用tess4j实现OCR的最简单样例
网上很多教程没有介绍清楚tessdata的位置,以及怎么配置,并且对中文库的描述也存在问题,这里介绍一个最简单的样例. 1.使用maven,直接引入依赖,确保你的工程JDK是1.8以上 <dep ...
- 【Selenium-WebDriver实战篇】基于java的selenium之验证码识别内容
==================================================================================================== ...
- 百度DMLC分布式深度机器学习开源项目(简称“深盟”)上线了如xgboost(速度快效果好的Boosting模型)、CXXNET(极致的C++深度学习库)、Minerva(高效灵活的并行深度学习引擎)以及Parameter Server(一小时训练600T数据)等产品,在语音识别、OCR识别、人脸识别以及计算效率提升上发布了多个成熟产品。
百度为何开源深度机器学习平台? 有一系列领先优势的百度却选择开源其深度机器学习平台,为何交底自己的核心技术?深思之下,却是在面对业界无奈时的远见之举. 5月20日,百度在github上开源了其 ...
- 基于Tesseract的OCR识别小程序
一.背景 先说下开发背景,今年有次搬家找房子(2020了应该叫去年了),发现每天都要对着各种租房广告打很多电话.(当然网上也找了实地也找),每次基本都是对着墙面看电话号码然后拨打,次数一多就感觉非常麻 ...
- 基于Tesseract组件的OCR识别
基于Tesseract组件的OCR识别 背景以及介绍 欲研究C#端如何进行图像的基本OCR识别,找到一款开源的OCR识别组件.该组件当前已经已经升级到了4.0版本.和传统的版本(3.x)比,4.0时代 ...
- Atitit ocr识别原理 与概论 attilax总结
Atitit ocr识别原理 与概论 attilax总结 1.1. Ocr的过程与流程1 1.2. OCR不同技术细分略有不同,但大概原理是一样的. 即主要技术过程是:二值化(又叫归一化)----- ...
- 基于Python实现对PDF文件的OCR识别
http://www.jb51.net/article/89955.htm https://pythontips.com/2016/02/25/ocr-on-pdf-files-using-pytho ...
- OCR识别-python版(一)
需求:识别图片中的文字信息环境:windows系统 开发语言:python 使用工具类:1.pyocr 2.PIL 3.tesseract-ocr 步骤: 1.pyocr 网络通直接使用命令:pip ...
- 汽车Vin码识别——可以嵌入到手机里的新OCR识别技术
汽车Vin码识别(车架号识别),顾名思义,就是识别汽车的Vin码(车架号),汽车Vin码识别(车架号识别)利用的是OCR识别技术,支持视频流获取图像,自动触发识别,另外汽车Vin码 ...
随机推荐
- svn already lock解决方法
svn在pull的时候,出现svn already lock 错误.只需要在Cleanup里面勾选Break locks
- cocos 3.16 生成预编译库模板
打开模板功能 不知道什么原因,cocos 3.16关闭了模板功能,因此使用cocos -h命令时,无法显示gen-templates参数. 打开的方式也挺简单,打开tools/cocos2d-cons ...
- mysql_test
------------------ #/bin/sh binlogfile=$1 if [ ! -n $binlogfile ]thenecho "pls input your mysql ...
- PHP 汉字数字互转(100以内)| 汉字转数字 | 数字转汉字
<?php function numDatabase(){ $numarr =array(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,2 ...
- 【git】之clone(克隆)
直接克隆 git clone https://github.com/gyjx/test.git 指定克隆某个分支 git clone -b dev https://github.com/gyjx/te ...
- 把1,2,3…n*n 的数字按照顺时针螺旋的形式填入数字矩阵
从键盘输入一个整数(1~20)则以该数字为矩阵的大小,把1,2,3…n*n 的数字按照顺时针螺旋的形式填入其中.例如:输入数字2,则程序输出:1 24 3输入数字3,则程序输出:1 2 38 9 47 ...
- RPM包安装软件 -- 详细解读
一.RPM包命名规则 1.RPM包在哪 RPM包在光盘中 2.RPM包命名原则 httpd-2.2.15-15.e16.centos.1.i686.rpm httpd 软件包名 2.2.15 软件版本 ...
- CSS自定义滚动条样式
原文地址:http://www.qianduan.net/css-custom-scroll-bar-style/ 相信很多人都遇到过在设计中自定义滚动条样式的情景,之前我都是努力说服设计师接受浏览器 ...
- jvm感知docker容器参数
docker中的jvm检测到的是宿主机的内存信息,它无法感知容器的资源上限,这样可能会导致意外的情况. -m参数用于限制容器使用内存的大小,超过大小时会被OOMKilled. -Xmx: 默认为物理 ...
- python:逻辑运算与编码
一. 1.pycharm的使用 2.in not in 的使用 in not in 为了查找数据中是否存在需要查找的数据, in如果存在返回True,不存在返回False (not i ...