Boundary-based Real-time Text Detection on Container Code

Liu, Kuikun; Sun, Cai; Chi, Haoyuan

doi:10.1109/iscsic54682.2021.00025

Cited by 4 publications

(1 citation statement)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The proposed approach can proceed at 1.13 FPS and has the accuracy of 93.33%. Liu et al [26] further improved the accuracy of the localization by a real-time ML system to predict the texts and their boundaries, consequently fuse the results to improve the segmentation accuracy. Their system achieved a F-measure result of 96.5% at 70 FPS while performed on the on-field datasets.…”

Section: Introductionmentioning

confidence: 99%

Digital transformation for shipping container terminals using automated container code recognition

2023

View full text Add to dashboard Cite

Due to the sweeping waves of global industry development, the number of containers passing through terminal ports increases every day. Therefore, it is essential to automate the identification process for the container codes to replace the manual identification for more efficient logistics and safer workplace. This paper aims to design and evaluate the performance of such a system. Specifically, automated container codes recognition (ACCR) has been implemented. This is a novel container tracking model based on image processing algorithms and machine learning (ML) algorithms to be applied in ports. There are three steps in this system: character detection, character isolation, and character recognition. The first step is to identify an area with 10 digits and 26 capitals. After detecting the text area, the second step is to separate the characters. Each character is recognized in the last step by the classification method. In particular, features are extracted with the histogram of oriented gradients (HOG) algorithm and support vector machines (SVMs) for training and prediction. The trained ML model is then used to classify characters and digits according to what it has learned. In general, the digital technologies in logistics and container management in ports will benefit from the proposed algorithms.

show abstract

Section: Introductionmentioning

confidence: 99%