Mirrors/Stirling-PDF
Mirrors/Stirling-PDF
1
0
Fork 0
mirror of https://github.com/Frooodle/Stirling-PDF.git synced 2025-09-12 17:52:13 +02:00
Code Issues Packages Projects Releases Wiki Activity

revert OCR

Browse Source
This commit is contained in:
Anthony Stirling 2025-03-25 13:35:34 +00:00
parent 1597d23f92
commit e13cb19439
1 changed files with 10 additions and 41 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
src/main/java/stirling/software/SPDF/controller/api/misc/OCRController.java
View File

@ -9,11 +9,7 @@ import java.util.stream.Collectors;
import java.util.zip.ZipEntry; import java.util.zip.ZipEntry;
import java.util.zip.ZipOutputStream; import java.util.zip.ZipOutputStream;
import javax.imageio.IIOImage;
import javax.imageio.ImageIO; import javax.imageio.ImageIO;
import javax.imageio.ImageWriteParam;
import javax.imageio.ImageWriter;
import javax.imageio.stream.FileImageOutputStream;
import org.apache.pdfbox.multipdf.PDFMergerUtility; import org.apache.pdfbox.multipdf.PDFMergerUtility;
import org.apache.pdfbox.pdmodel.PDDocument; import org.apache.pdfbox.pdmodel.PDDocument;
@ -92,6 +88,7 @@ public class OCRController {
Files.createDirectories(tempImagesDir); Files.createDirectories(tempImagesDir);
Process process = null; Process process = null;
try { try {
// Save input file
inputFile.transferTo(tempInputFile.toFile()); inputFile.transferTo(tempInputFile.toFile());
PDFMergerUtility merger = new PDFMergerUtility(); PDFMergerUtility merger = new PDFMergerUtility();
merger.setDestinationFileName(finalOutputFile.toString()); merger.setDestinationFileName(finalOutputFile.toString());
@ -101,6 +98,7 @@ public class OCRController {
for (int pageNum = 0; pageNum < pageCount; pageNum++) { for (int pageNum = 0; pageNum < pageCount; pageNum++) {
PDPage page = document.getPage(pageNum); PDPage page = document.getPage(pageNum);
boolean hasText = false; boolean hasText = false;
// Check for existing text
try (PDDocument tempDoc = new PDDocument()) { try (PDDocument tempDoc = new PDDocument()) {
tempDoc.addPage(page); tempDoc.addPage(page);
PDFTextStripper stripper = new PDFTextStripper(); PDFTextStripper stripper = new PDFTextStripper();
@ -115,42 +113,12 @@ public class OCRController {
Path pageOutputPath = Path pageOutputPath =
tempOutputDir.resolve(String.format("page_%d.pdf", pageNum)); tempOutputDir.resolve(String.format("page_%d.pdf", pageNum));
if (shouldOcr) { if (shouldOcr) {
// Render with lower DPI (200 instead of 300) // Convert page to image
BufferedImage image = pdfRenderer.renderImageWithDPI(pageNum, 200); BufferedImage image = pdfRenderer.renderImageWithDPI(pageNum, 300);
Path imagePath =
// Convert to RGB to remove alpha channel if present tempImagesDir.resolve(String.format("page_%d.png", pageNum));
if (image.getType() != BufferedImage.TYPE_INT_RGB) { ImageIO.write(image, "png", imagePath.toFile());
BufferedImage rgbImage = // Build OCR command
new BufferedImage(
image.getWidth(),
image.getHeight(),
BufferedImage.TYPE_INT_RGB);
rgbImage.getGraphics().drawImage(image, 0, 0, null);
image = rgbImage;
}
// Save as JPEG with compression
String imageName = String.format("page_%d.jpg", pageNum);
Path imagePath = tempImagesDir.resolve(imageName);
Iterator<ImageWriter> writers = ImageIO.getImageWritersByFormatName("jpg");
if (!writers.hasNext()) {
throw new IllegalStateException("No JPG ImageWriter found");
}
ImageWriter writer = writers.next();
ImageWriteParam params = writer.getDefaultWriteParam();
params.setCompressionMode(ImageWriteParam.MODE_EXPLICIT);
params.setCompressionQuality(0.7f); // Adjust quality here (0.7 = 70%)
try (FileImageOutputStream output =
new FileImageOutputStream(imagePath.toFile())) {
writer.setOutput(output);
writer.write(null, new IIOImage(image, null, null), params);
} finally {
writer.dispose();
}
// Build OCR command with JPG image
List<String> command = new ArrayList<>(); List<String> command = new ArrayList<>();
command.add("tesseract"); command.add("tesseract");
command.add(imagePath.toString()); command.add(imagePath.toString());
@ -160,6 +128,7 @@ public class OCRController {
.toString()); .toString());
command.add("-l"); command.add("-l");
command.add(String.join("+", languages)); command.add(String.join("+", languages));
// Always output PDF
command.add("pdf"); command.add("pdf");
ProcessBuilder pb = new ProcessBuilder(command); ProcessBuilder pb = new ProcessBuilder(command);
process = pb.start(); process = pb.start();