Modelling Studies of Automatic Container Code Recognition System for Real Time Implementation

The need for an industrial competent Automatic Container Code Recognition (ACCR) has reached an unprecedented high due to the revolution of container shipping. The productivity of internal trades relies on an efficient and effective systematic container management plan, and the increase in that has resulted in the demand for a proficient ACCR system. In the recent development of deep learning, the emerging deep learning models not only addressed the interferences of irrelevant words in natural scenes that are highly challenging in traditional approaches but resulted in a simpler yet elegant pipeline. This paper proposes a practical end-to-end, real-time deployable ACCR system that runs at an inference speed of 31.72 frames per second (fps) with an end-to-end accuracy of 89.20%. The ACCR system is composed of localization and recognition modules that implement Differential Binarization (DB) and Semantic Reasoning Network (SRN) respectively to address the Container Code Character (CCC) in Scene Text Recognition (STR). Several other neural networks such as Single-Shot Arbitrarily Shaped Text Detector (SAST) and Convolutional Recurrent Neural Network (CRNN) with Connectionist Temporal Classification (CTC) feature have been studied and investigated to identify their inherent strength and weakness in addressing STR problems. Ultimately, a graphical user interface tool was developed for practical engineering design. Code is available at: https://github.com/laulj/CCCR.