tess4j的簡單使用doOCR

Warning: Parameter not found: enable_new_segsearch
Warning: Parameter not found: save_raw_choices
D:\devtools\Tesseract-OCR\tessdata
result:--->650 3428
           

一、測試圖檔777.png

二、測試代碼：

package com.gazgeek.helloworld.tess4jTest;


import java.awt.*;
import java.io.File;
import net.sourceforge.tess4j.*;

public class Testtess {

    public static void main(String[] args) {

        File imageFile = new File("F:\\imgall\\777.png");
        Tesseract tessInst = new Tesseract();
        tessInst.setDatapath("D:\\devtools\\Tesseract-OCR\\tessdata");
        tessInst.setLanguage("eng");// eng.traineddata is in /tessdata direcotry

        try {
            String result= tessInst.doOCR(imageFile);
            System.out.println("D:\\devtools\\Tesseract-OCR\\tessdata");
            System.out.println("result:--->"  + result );
        } catch (TesseractException e) {
            System.err.println(e.getMessage());
        }

    }

}
           

三、測試結果：

四、FAQ

1. ERROR net.sourceforge.tess4j.Tesseract - Not a JPEG file: starts with 0x89 0x50

Solution: the file is not acctually JPEG file, select true JPEG file.

2. WARN Please make sure the TESSDATA_PREFIX environment variable is set to your "tessdata" directory.

Solution: option A, tessInst.setDatapath(System.getProperty("user.dir") + "/tessdata");

               option B, set TESSDATA_PREFIX your environment. Which is Tesseract's tessdata default value. If do not set, it will

               open ./*.traineddata file.

3. "Warning: Parameter not found: enable_new_segsearch" 

Solution: Works with this eng.traineddata: https://github.com/tesseract-ocr/tessdata_fast/blob/master/eng.traineddata

Note: language data file best use tessdata_best's file. If you want to recognize chinese, select chi_sim.traineddata, and download it, move it in your tessdata directory.

Java's print API basically works on the assumption that everything is done at 72 dpi. This means that you can use this as bases for converting to/from different measurements

references:

1. http://www.jbrandsma.com/news/2015/12/07/ocr-with-java-and-tesseract/

2. https://sourceforge.net/projects/tess4j/

3. https://github.com/tesseract-ocr/tessdata_best

4. https://www.b4x.com/android/forum/threads/solved-tesseract-api-a-120-opotunity.101482/

6. https://www.learnopencv.com/deep-learning-based-text-recognition-ocr-using-tesseract-and-opencv/

7. https://stackoverflow.com/questions/18975595/how-to-design-an-image-in-java-to-be-printed-on-a-300-dpi-printer

五、《Mac安裝Tesseract的全過程，附帶完整的錯誤和異常的解決辦法。Java開源OCR識别》

MacOS

使用 

sudo port install tesseract

安裝 tesseract-ocr 之後，再執行以下指令時：

sudo tesseract 1563327696899809.jpg 1563327696899809.txt -l chi_sim+eng

提示：

Warning: Parameter not found: enable_new_segsearch

Tesseract Open Source OCR Engine v4.0.0 with Leptonica

Warning: Invalid resolution 0 dpi. Using 70 instead.

Estimating resolution as 194

這是因為 -l 參數中指定的語言包都沒有安裝的原因，使用以下指令安裝：

sudo port install tesseract-chi-sim

sudo port install tesseract-eng

異常彙總：

①Warning: Parameter not found: enable_new_segsearch

Mac出現的時候（把語言封包件拷貝到你在Java代碼設定好的目錄下，原因是此目錄沒有中文簡體的語言包）

1. ITesseract iTesseract = new Tesseract();

2. iTesseract.setDatapath("你的語言包絕對路徑");

②Warning: Invalid resolution 0 dpi. Using 70  instead.

1. ITesseract iTesseract = new Tesseract();

2. iTesseract.setDatapath("你的語言包絕對路徑");

3. iTesseract.setTessVariable("user_defined_dpi", "300");

設定一下dpi即可，預設設定300是最好的