Finding word bounding boxes

Aspose.OCR allows you to automatically find the coordinates of image rectangles containing individual words in text. This can be useful for highlighting detected words when previewing an image.

To get bounding boxes of all words of the image, use getTextAreas method of AsposeOCR class. Specify AreasType.WORDS as the areaType parameter of the method. isDetectAreas parameter of the method is ignored.

This method works for images in the following formats: GIF, PNG, JPEG, BMP, WBMP.

From file

AsposeOCR api = new AsposeOCR();
ArrayList<Rectangle> paragraphs = api.getTextAreas("source.png", AreasType.WORDS);
paragraphs.forEach((region) -> {
	System.out.println(region.recognitionText);
});

From stream

AsposeOCR api = new AsposeOCR();
BufferedImage source = ImageIO.read(new File("source.png"));
ArrayList<Rectangle> paragraphs = api.getTextAreas(source, AreasType.WORDS);
paragraphs.forEach((region) -> {
	System.out.println(region.recognitionText);
});

Coordinates of each word (top-left corner, width and height) are returned as a list of Rectangle objects.