Open CV-Based Analogue Gauge Detection and Digitization: A Sturdy, Computationally Effective Framework for Industrial Retrofitting

Abstract

Despite the rapid proliferation of Industry 4.0 and the Industrial Internet of Things (IIoT), analogue gauges remain deeply entrenched in critical industrial infrastructure due to their inherent robustness, passive operation, and zero electrical dependency. Their continued use, however, creates significant data silos, impeding real-time monitoring, predictive maintenance, and operational efficiency. This paper presents a comprehensive computer vision framework for the automated digitization of circular analogue gauges using the OpenCV library. The proposed pipeline employs the Hough Circle Transform (HCT) for gauge localization and the Probabilistic Hough Line Transform (PHLT) for pointer segmentation, augmented by adaptive image preprocessing and angular interpolation for value computation. Unlike data-hungry deep learning architectures requiring specialized GPU hardware, the proposed classical approach employs a semi-automated human-in-the-loop calibration strategy that ensures generalizability across diverse gauge types and configurations. Experimental evaluation demonstrates realtime performance at approximately 28 frames per second on a Raspberry Pi 4 Model B (4 GB RAM) with a mean reading error between 0.1% and 0.5%. The system achieves competitive accuracy relative to state-of-the-art deep learning methods while operating at a fraction of the computational and financial cost, making it a viable solution for retrofitting legacy industrial facilities