/**
* Test of doOCR method, of class Tesseract.
*
* @throws Exception while processing image.
*/
@Test
public void testDoOCR_List_Rectangle() throws Exception {
File imageFile = null;
String expResult = "The (quick) [brown] {fox} jumps!\nOver the $43,456.78 <lazy> #90 dog";
String result = "<empty>";
try {
logger.info("doOCR on a PDF document");
imageFile = new File(this.testResourcesDataPath, "eurotext.pdf");
List<IIOImage> imageList = ImageIOHelper.getIIOImageList(imageFile);
result = instance.doOCR(imageList, null);
logger.info(result);
assertEquals(expResult, result.substring(0, expResult.length()));
} catch (IOException e) {
logger.error(
"Exception-Message: '{}'. Imagefile: '{}'",
e.getMessage(),
imageFile.getAbsoluteFile(),
e);
fail();
} catch (TesseractException e) {
logger.error(
"Exception-Message: '{}'. Imagefile: '{}'",
e.getMessage(),
imageFile.getAbsoluteFile(),
e);
fail();
} catch (StringIndexOutOfBoundsException e) {
logger.error(
"Exception-Message: '{}'. Imagefile: '{}'",
e.getMessage(),
imageFile.getAbsoluteFile(),
e);
fail();
}
}
/**
* Test of doOCR method, of class Tesseract.
*
* @throws Exception while processing image.
*/
@Test
public void testDoOCR_File() throws Exception {
logger.info("doOCR on a PNG image");
File imageFile = new File(this.testResourcesDataPath, "eurotext.png");
String expResult = "The (quick) [brown] {fox} jumps!\nOver the $43,456.78 <lazy> #90 dog";
String result = instance.doOCR(imageFile);
logger.info(result);
assertEquals(expResult, result.substring(0, expResult.length()));
}
/**
* Test of doOCR method, of class Tesseract.
*
* @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, "eurotext.bmp");
Rectangle rect =
new Rectangle(
0, 0, 1024, 800); // define an equal or smaller region of interest on the image
String expResult = "The (quick) [brown] {fox} jumps!\nOver the $43,456.78 <lazy> #90 dog";
String result = instance.doOCR(imageFile, rect);
logger.info(result);
assertEquals(expResult, result.substring(0, expResult.length()));
}
/**
* Test of doOCR method, of class Tesseract.
*
* @throws java.lang.Exception
*/
@Test
public void testDoOCR_File_With_Configs() throws Exception {
logger.info("doOCR with configs");
File imageFile = new File(this.testResourcesDataPath, "eurotext.png");
String expResult = "[-0123456789.\n ]+";
List<String> configs = Arrays.asList("digits");
instance.setConfigs(configs);
String result = instance.doOCR(imageFile);
logger.info(result);
assertTrue(result.matches(expResult));
instance.setConfigs(
null); // since Tesseract instance is a singleton, clear configs so the effects do not carry
// on into subsequent runs.
}
/**
* Test of doOCR method, of class Tesseract.
*
* @throws java.lang.Exception
*/
@Test
public void testDoOCR_UNLV_Zone_File() throws Exception {
logger.info("doOCR on a PNG image with UNLV zone file .uzn");
// UNLV zone format: left top width height label
String filename = String.format("%s/%s", this.testResourcesDataPath, "eurotext_unlv.png");
File imageFile = new File(filename);
String expResult =
"& duck/goose, as 12.5% of E-mail\n\n"
+ "from [email protected] is spam.\n\n"
+ "The (quick) [brown] {fox} jumps!\n"
+ "Over the $43,456.78 <lazy> #90 dog";
String result = instance.doOCR(imageFile);
logger.info(result);
assertEquals(expResult, result.trim());
}
/**
* Test of deskew algorithm.
*
* @throws Exception while processing image.
*/
@Test
public void testDoOCR_SkewedImage() throws Exception {
logger.info("doOCR on a skewed PNG image");
File imageFile = new File(this.testResourcesDataPath, "eurotext_deskew.png");
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 expResult = "The (quick) [brown] {fox} jumps!\nOver the $43,456.78 <lazy> #90 dog";
String result = instance.doOCR(bi);
logger.info(result);
assertEquals(expResult, result.substring(0, expResult.length()));
}
@Test
public void testFolderExtraction() {
File tessDataFolder = null;
try {
/** Loads the image from resources. */
String filename = String.format("%s/%s", "/test-data", "eurotext.pdf");
URL defaultImage = getClass().getResource(filename);
File imageFile = new File(defaultImage.toURI());
/** Extracts <code>tessdata</code> folder into a temp folder. */
logger.info("Loading the tessdata folder into a temporary folder.");
tessDataFolder = LoadLibs.extractTessResources("tessdata");
/** Gets tesseract instance and sets data path. */
ITesseract instance = new Tesseract();
if (tessDataFolder != null) {
logger.info(tessDataFolder.getAbsolutePath());
instance.setDatapath(tessDataFolder.getAbsolutePath());
instance.setLanguage("por");
}
/** Performs OCR on the image. */
String result = instance.doOCR(imageFile);
logger.info(result);
} catch (TesseractException e) {
logger.error(e.getMessage());
logger.error(e.getMessage(), e);
} catch (URISyntaxException e) {
logger.error(e.getMessage(), e);
}
// checks if tessdata folder exists
assertTrue(tessDataFolder != null && tessDataFolder.exists());
}