A Highly Efficient Neural Network Solution for Automated Detection of Pointer Meters with Different Analog Scales Operating in Different Conditions

Abstract

<jats:p>We investigate a neural network–based solution for the Automatic Meter Reading detection problem, applied to analog dial gauges. We employ a convolutional neural network with a non-linear Network in Network kernel. Presently, there is a significant interest in systems for automatic detection of analog dial gauges, particularly in the energy and household sectors, but the problem is not yet sufficiently addressed in research. Our method is a universal three-level model that takes an image as an input and outputs circular bounding areas, object classes, grids of reference points for all symbols on the front panel of the device and positions of display pointers. Since all analog pointer meters have a common nature, this multi-cascade model can serve various types of devices if its capacity is sufficient. The model is using global regression for locations of symbols, which provides resilient results even for low image quality and overlapping symbols. In this work, we do not focus on the pointer location detection since it heavily depends on the shape of the pointer. We prepare training data and benchmark the algorithm with our own framework a3net, not relying on third-party neural network solutions. The experimental results demonstrate the versatility of the proposed methods, high accuracy, and resilience of reference points detection.</jats:p>

Dynamic Acoustic Unit Augmentation with BPE-Dropout for Low-Resource End-to-End Speech Recognition

“…The main end-to-end model architecture for our experiments was the Transformer. The encoder consisted of a 2-layer CONV2D subsampling block (to reduce input feature sequence by four times) (In object detection, Convolutional Neural Networks (CNNs) are used as main architecture blocks (e.g., [ 41 ]). Some inference-efficient ASR systems also use purely-convolutional solutions (e.g., [ 30 ]).…”

Real-time analogue gauge transcription on mobile phone

“…The task of analogue gauge transcription has been tackled numerous times. Recent methods fall largely into two groups: those which use traditional based computer vision [3,20,25,8,18,23,9,26,17] and those which utilise deep learning [14,10,11,21,1,16,5,15]. Methods using traditional based approaches are typically brittle to appearance variation in lighting, background clutter and highly constrained to particular types of gauges.…”

A High-Precision Automatic Pointer Meter Reading System in Low-Light Environment

“…Fang et al [ 6 ] proposed detecting meters using Mask R-CNN [ 7 ] to detect needle and scale key points and calculate readings by semantic segmentation. Alexeev [ 8 ] proposed a NiN-based AMR detector to detect pointer meters with different analog ranges, automatically. Salomon et al [ 9 ] detected the bounding box of the dial by YOLO-based methods, and Faster-RCNN extracted the needle and then calculated the readings.…”