Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : Media of Computer Science

Comparative Analysis of Google Vision OCR with Tesseract on Newspaper Text Recognition Prakisya, Nurcahya Pradana Taufik; Kusmanto, Bintang Timur; Hatta, Puspanda
Media of Computer Science Vol. 1 No. 1 (2024): June 2024
Publisher : CV. Digital Innovation

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.69616/mcs.v1i1.178

Abstract

Optical Character Recognition (OCR) is a technique used to convert image files into machine-readable text. There are two Optical Character Recognition (OCR) algorithms that are currently well known and widely used, namely Google Vision's Optical Character Recognition (OCR) and Tesseract. The purpose of this study is to compare the Optical Character Recognition (OCR) algorithms of Google Vision and Tesseract so that people can more easily find out which algorithm is the right one to implement on the system they are going to build. The method used in this research is Research and Development (R&D) with the stages of literature study, needs analysis, dataset collection and expansion, architectural design development and application modeling, system implementation, testing and evaluation, drawing conclusions. To be able to determine the level of accuracy, precision and sensitivity of each algorithm, this research uses the Confusion Matrix formula. The results of this study conclude that Google Vision's Optical Character Recognition (OCR) algorithm is superior to Tesseract because the level of accuracy, sensitivity, and precision is superior to Google Vision.
Preprocessing Image for License Plate Detection: A Systematic Literature Review Prasetyo, Riyan Bagas Dwi; Abdullayev, Vugar; Prakisya, Nurcahya Pradana Taufik; Sujana, Yudianto; Siswanto, Rahmat
Media of Computer Science Vol. 2 No. 2 (2025): December 2025
Publisher : CV. Digital Innovation

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.69616/mcs.v2i2.241

Abstract

Rapid population growth contributes to an increase in the volume of vehicles, creating major challenges in their management. One potential solution is the application of deep learning-based artificial intelligence technology for automatic detection of vehicle license plates. This research uses a Systematic Literature Review (SLR) approach to evaluate the performance of various deep learning architectures in the detection process. Out of 125 articles identified, 20 articles were selected based on specific selection criteria. The analysis revealed that preprocessing techniques, such as HE, AHE, ECHE, CLAHE, and ECLACHE, have significant contributions in the processing of vehicle license plate datasets. These techniques were able to improve the visual quality of the images, thus supporting the detection process with an accuracy rate of more than 95%. This research also identified challenges, such as high computational requirements and large-scale data processing. Further research is recommended to apply preprocessing on standardized datasets to develop a reliable, efficient and sustainable detection system.