Convert images and PDFs to text in C++

Extracts text from images and creates searchable PDFs

Supports any image you can get form a scanner or camera

Reads Extended Latin and Cyrillic scripts

Recognizes over 6,000 Chinese characters

Detects and recognizes all popular typefaces and formatting

Pre-processes images before recognition

Processes the whole image or selected areas only

Supports rotated, skewed and noisy images

Batch recognition of all images in a folder or archive

Recognizes images provided as web links

Finds and automatically corrects misspelled words

Get recognition result as JSON

Easy to Use

Do you still think C++ OCR is hard? With our library, you only need 5 lines of code to recognize the image and display the result. Try this code and see for yourself:

// Prepare buffer for result
const size_t len = 4096;
const buffer[len] = { 0 };
// Do the magic
size_t size = aspose::ocr::page(L"<file name>", buffer, len);
// Display the recognition result
std::wcout << buffer << L"\n";

26 Recognition Languages

OCR API can recognize a large number of languages and all popular writing scripts, including texts with mixed languages.

• Extended Latin alphabet: Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Italian, Latvian, Lithuanian, Norwegian, Polish, Portuguese, Romanian, Slovak, Slovenian, Spanish, Swedish.
• Cyrillic alphabet: Belorussian, Bulgarian, Kazakh, Russian, Serbian, Ukrainian.
• Chinese: more than 6,000 characters.

You can leave the language detection to the library or define the language yourself to increase the recognition performance and reliability.

Batch Processing

OCR API frees you from recognizing every image one-by-one by offering various batch-processing methods that allow you to recognize multiple images in one call:

• Recognition of multi-page PDF and TIFF files.
• Recognition of all files in a folder.
• Recognition of all files in an archive.

// Provide archive path
std::string archive_path = "book.zip";

// Prepare buffer for result
const size_t len = 4096;
wchar_t buffer[len] = { 0 };

// Initialize RecognitionSettings object with default values
RecognitionSettings settings;

// recognize
size_t res_len = aspose::ocr::pages_multi(archive_path.c_str(), buffer, len, settings);

Preserve Formatting

The OCR library reads all popular typefaces such as Arial, Times New Roman, Courier New, Tahoma, Calibri and more in regular, bold and italic styles and carefully preserves formatting in OCR results. You can also split recognition results into lines and detect text areas in a page.

Recognize Photos

The widespread adoption of OCR applications is usually stopped by the fact that scanners are not commonplace for most users. Our OCR library has powerful built-in image pre-processing filters that can handle dark, rotated, skewed, and noisy images. In combination with support for all image formats, it allows for reliable recognition of even smartphone photos. Most of the pre-processing and image correction is done automatically, so you will only have to intervene in difficult cases.

// Original image
std::string image_path = "../Data/Source/sample.png";
rect rectangles[2] = { {90, 186, 775, 95} , { 928, 606, 790, 160 } };

// Prepare buffer for result
const size_t len = 4096;
wchar_t buffer[len] = { 0 };

// Adjust skew angle
RecognitionSettings settings;
settings.format = export_format::text;
settings.rectangles = rectangles;
settings.rectangles_size = 2;
settings.skew = 5;

// Recognize image
size_t res_len = aspose::ocr::page_settings(image_path.c_str(), buffer, len, settings);

Spell Check

While the OCR produces reliable results, dust and print defects might cause some symbols to be recognized incorrectly. OCR API has a built-in spell checker that automatically replaces misspelled words and frees you from having to manually correct the recognition results.