ID Card Number Recognition Project Using JavaCV, OpenCV, and Tess4J

The project combines the open-source ID card OCR repository with Tess4J, allowing direct use without training, while also supporting optional training. It replaces the original OpenCV installation process by using JavaCPP to embed required C++ libraries, eliminating the need for a separate OpenCV setup. A new front‑end feature lets users select the recognition area, and the back‑end validates the extracted data. Images are optimized and cropped before OCR using Tess4J to recognize digits and the character "x".

Encountered Issue:

java.lang.UnsatisfiedLinkError caused by missing C++ runtime libraries; resolved by installing the 64‑bit Visual C++ Redistributable (vc_redist.x64.exe).

ID Card Number Recognition:

The service runs at http://localhost:8080/idCard/index , built on OpenCV and Java, offering high recognition rates (over 90% accuracy on clear images).

Required Software

Windows 7 64‑bit

JDK 1.8.0_45

JUnit 4

OpenCV 4.3

JavaCPP 1.5.3

Tess4J 4.5.1

Tesseract 4.0.0

Project Updates:

Replaced base64 image upload with WebUploader chunked upload to improve speed, especially on paid browsers; page renamed to idcard_bak.html .

Unified test image save paths into a configuration document.

Upgraded OpenCV from 3.4.3 to 4.3.

Project Repository: https://gitee.com/endlesshh/idCardCv

PS: If you find this sharing helpful, feel free to like and view.