Estimating One-Dimensional Barcode Image Orientation by Using 2D Fourier Transform: A CPU-Based Approach

Abstract

Barcode detection and decoding are critical tasks in numerous industrial sectors, including inventory management, logistics, and retail. Despite the advancements in Artificial Intelligence (AI) and Machine Learning (ML), many existing barcode detection methods rely heavily on GPU-accelerated techniques or are confined by specific angle requirements, limiting their versatility and computational efficiency. In this paper, a novel approach for stabilizing barcode image orientations using the 2D Fourier Transform was developed, with an emphasis on CPU-based implementation.