Search code examples
javamavenjakarta-eetesseracttess4j

RuntimeException when trying to use Tess4J in Java EE

Im trying to use Tess4J in Java EE (Payara server), is this possible and if so how?

Exact Exception I'm getting:

e = (net.sourceforge.tess4j.TesseractException) net.sourceforge.tess4j.TesseractException: java.lang.RuntimeException: Need to install JAI Image I/O package. https://java.net/projects/jai-imageio/

I have added the jai-imageio to my pom.xml, as well as added it to the modules of Payara.

File pom.xml

    <!-- https://mvnrepository.com/artifact/net.sourceforge.tess4j/tess4j -->
    <dependency>
        <groupId>net.sourceforge.tess4j</groupId>
        <artifactId>tess4j</artifactId>
        <version>3.4.1</version> <!-- used 3.4.2 as well -->
    </dependency>

    <!-- https://mvnrepository.com/artifact/com.github.jai-imageio/jai-imageio-core -->
    <dependency>
        <groupId>com.github.jai-imageio</groupId>
        <artifactId>jai-imageio-core</artifactId>
        <version>1.3.1</version>
        <scope>runtime</scope>  <!-- tried without this as well -->
    </dependency>

Added JAR to 

`Payara\glassfish\modules`

Tess4J code (If any improvements can be made to this as well it would be appreciated).

       ITesseract instance = new Tesseract();
        instance.setDatapath(pLangaugePath); // C:\\t
        instance.setLanguage(pLanguage); // eng

            try {
                File[] tifFiles = PdfUtilities.convertPdf2Png(pFile);

                if (tifFiles != null) {

                    for (File tifFile : tifFiles) {
                        String ocrText = instance.doOCR(tifFile);

                        if (StringUtils.isNotBlank(ocrText)) {
                            ret.append(ocrText);
                        }
                    }
                }
            } catch (TesseractException e) {
                LOG.error("Could not do ocr on image file created via pdf ", e);
            }

Have tried the following 2 examples as well. 1.

     try (PDDocument document = PDDocument.load(pFile)) {
                int totalPages = document.getNumberOfPages();

                PDFRenderer renderer = new PDFRenderer(document);

                for (int pi = 0; pi < totalPages; pi++) {
                    BufferedImage image = renderer.renderImageWithDPI(pi, 75);

                    String ocrText = instance.doOCR(image);

                    if (StringUtils.isNotBlank(ocrText)) {
                        ret.append(ocrText);
                    }
                }
            } catch (Exception e) {
                LOG.error("Could not do ocr on pdf", e);
            }

2.

 try {

        ITesseract instance = new Tesseract();
        instance.setDatapath(pLangaugePath); // C:\\t
        instance.setLanguage(pLanguage); // eng

        String ocrText = instance.doOCR(pFile);

        if (StringUtils.isNotBlank(ocrText)) {
            ret.append(ocrText);
        }

    } catch (Exception e) {
        LOG.error("Could not do ocr on image file created via pdf ", e);
    }

Research:

Found this Didnt work / solution

as well as didnt work

Solution

  • Tess4J was known for not working with Glassfish due to the run-time exception caused by the unavailability of JNA RESOURCE_PREFIX string constant. This issue has been fixed in the latest releases 3.4.9 (for Tesseract 3.05.01) and 4.0.2 (for Tesseract 4.0.0-beta.1). The library can now be used with Glassfish, and perhaps Payara Server.

    You may also need to include ImageIO.scanForPlugins(); statement before the OCR call. That is meant to ensure the appropriate ImageReader be available to read input images.