Design and Implementation of Lightweight Vehicle License Plate Recognition Module utilizing OpenCV and Tesseract OCR Library

Background/Objectives: In order to recognize the license plates automatically, we design and implement a vehicle license plate recognition module that extracts characters of license plate area using open source OpenCV and Terreract OCR library. Methods/Statistical analysis: The static image was binarized using OpenCV 's binarization function. After binarizing the image by adjusting the pixel values between adjacent pixels, the candidate region judged to be a license plate was derived. The final candidate was derived according to the proposed algorithm in the candidate region. The extracted plate area was analyzed by using the Tesseract OCR library, and characters were extracted as a character string. Findings: The vehicle license plate recognition module relates to character recognition in the field of computer vision. In this paper, we designed and implemented a module that recognizes a license plate by using open source, applying a proposed algorithm to a moving object as a static image. The proposed module is a relatively lightweight software module and can be used in other applications. It is possible to install the camera at the entrance of the apartment and can read the license plate to identify whether it is a resident or not. When speeding and traffic violations occur on the highway, the vehicle numbers can be automatically stored and managed in the database. In addition, there is an advantage that it can be applied to various character recognition applications through modification of a slight algorithm in the module. Improvements/Applications: In addition to character recognition, the OpenCV library can be applied to various fields such as pattern recognition, object tracking, and motion recognition. Therefore, we will be able to create technologies corresponding to various services that are becoming automated and unmanned.