Zero-Shot Recognition of Utility Meter Readings Using Vision-Language Models and General-Purpose OCRs
Resumo
Supervised Automatic Meter Reading (AMR) methods achieve high accuracy but require costly domain-specific training. Despite the popularity of Vision-Language Models and general Optical Character Recognition systems offering theoretical zero-shot capabilities, their effectiveness on constrained real-world meters remains unexplored. We evaluated the zero-shot performance of seven models (PyTesseract, Keras-OCR, EasyOCR, PARSeq, GOT-OCR 2.0, DeepSeek-VL2 and Qwen2.5-VL) on the UFPR-AMR and Copel-AMR datasets. Results show supervised recognition rates average 95.87%, outperforming the best zero-shot model, which averaged 64.86% across both datasets. This accuracy gap makes generalist models currently unsuitable for AMR.Referências
Awais, M., Naseer, M., Khan, S., Anwer, R. M., Cholakkal, H., Shah, M., Yang, M.-H., and Khan, F. S. (2025). Foundation models defining a new era in vision: a survey and outlook. IEEE Transactions on Pattern Analysis and Machine Intelligence, 47(4):2245–2264.
Baek, J., Kim, G., Lee, J., Park, S., Han, D., Yun, S., Oh, S. J., and Lee, H. (2019). What is wrong with scene text recognition model comparisons? dataset and model analysis. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 4715–4723.
Bai, S., Chen, K., Liu, X., Wang, J., Ge, W., Song, S., Dang, K., Wang, P., Wang, S., Tang, J., Zhong, H., Zhu, Y., Yang, M., Li, Z., Wan, J., Wang, P., Ding, W., Fu, Z., Xu, Y., Ye, J., Zhang, X., Xie, T., Cheng, Z., Zhang, H., Yang, Z., Xu, H., and Lin, J. (2025). Qwen2.5-VL technical report.
Bautista, D. and Atienza, R. (2022). Scene text recognition with permuted autoregressive sequence models. In European Conference on Computer Vision, pages 178–196. Springer.
Bochkovskiy, A., Wang, C.-Y., and Liao, H.-Y. M. (2020). YOLOv4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934.
Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., et al. (2022). On the opportunities and risks of foundation models. arXiv preprint arXiv:2108.07258.
Calefati, A., Gallo, I., and Nawaz, S. (2019). Reading meter numbers in the wild. In 2019 Digital Image Computing: Techniques and Applications (DICTA), pages 1–6. IEEE.
Carvalho, R., Melo, J., Graça, R., Santos, G., and Vasconcelos, M. J. M. (2023). Deep learning-powered system for real-time digital meter reading on edge devices. Applied Sciences, 13(4):2315.
Gómez, L., Rusinol, M., and Karatzas, D. (2018). Cutting sayre’s knot: reading scene text without segmentation. application to utility meters. In 2018 13th IAPR International Workshop on Document Analysis Systems (DAS), pages 97–102. IEEE.
Gu, Z., Zhu, B., Zhu, G., Chen, Y., Tang, M., and Wang, J. (2024). AnomalyGPT: Detecting industrial anomalies using large vision-language models. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 1932–1940.
JaidedAI (2020). EasyOCR: Ready-to-use OCR. [link].
Laroca, R., Araujo, A. B., Zanlorensi, L. A., De Almeida, E. C., and Menotti, D. (2021). Towards image-based automatic meter reading in unconstrained scenarios: A robust and efficient approach. IEEE Access, 9:67569–67584.
Laroca, R., Barroso, V., Diniz, M. A., Gonçalves, G. R., Schwartz, W. R., and Menotti, D. (2019). Convolutional neural networks for automatic meter reading. Journal of Electronic Imaging, 28(1):013023–013023.
Levenshtein, V. I. et al. (1966). Binary codes capable of correcting deletions, insertions, and reversals. In Soviet Physics Doklady, volume 10, pages 707–710. Soviet Union.
Liao, S., Zhou, P., Wang, L., and Su, S. (2019). Reading digital numbers of water meter with deep learning based object detector. In Chinese Conference on Pattern Recognition and Computer Vision (PRCV), pages 38–49. Springer.
Morales, F. (2020). Keras-OCR: A packaged and flexible version of the CRAFT text detector and keras CRNN recognition model. [link].
Mueez, A., Baranwal, A., Chaj-Mejia, J., Bhatia, G., Voelker, J. T., and Vyas, S. (2026). Vision-language models for analog gauge reading: An empirical study of specialization, transfer and reliability. arXiv preprint arXiv:2608.17723.
Rodríguez, M., Berdugo, G., Jabba, D., Calle, M., and Jimeno, M. (2014). Hd-mr: A new algorithm for number recognition in electrical meters. Turkish Journal of Electrical Engineering and Computer Sciences, 22(1):87–96.
Shi, B., Bai, X., and Yao, C. (2016). An end-to-end trainable neural network for image-based sequence recognition and its application to scene text recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(11):2298–2304.
Shu, D., Ma, S., and Jing, C. (2007). Study of the automatic reading of watt meter based on image processing technology. In 2007 2nd IEEE Conference on Industrial Electronics and Applications, pages 2214–2217.
Smith, R. (2007). An overview of the tesseract OCR engine. In Ninth International Conference on Document Analysis and Recognition (ICDAR 2007), volume 2, pages 629–633. IEEE.
Spichkova, M., Van Zyl, J., Sachdev, S., Bhardwaj, A., and Desai, N. (2019). Easy mobile meter reading for non-smart meters: Comparison of aws rekognition and google cloud vision approaches. arXiv preprint arXiv:1910.12617.
Tong, R., Liu, J., Wang, T., Hu, X., Liu, S., Wang, L., and Xu, J. (2026). Does bigger mean better? comparative analysis of CNNs and biomedical vision-language models in medical diagnosis. In 2026 International Conference on Artificial Intelligence, Computer, Data Sciences and Applications (ACDSA), pages 1–6.
Vanetti, M., Gallo, I., and Nodari, A. (2013). Gas meter reading from real world images using a multi-net system. Pattern Recognition Letters, 34(5):519–526.
Wang, F., Weng, C., Wang, X., Chen, Z., Zhao, Z., and Tang, J. (2025). Dialbench: Towards accurate reading recognition of pointer meter using large foundation models. arXiv preprint arXiv:2511.21982.
Wang, Y., Chen, Q., Hong, T., and Kang, C. (2019). Review of smart meter data analytics: Applications, methodologies, and challenges. IEEE Transactions on Smart Grid, 10(3):3125–3148.
Wei, H., Liu, C., Chen, J., Wang, J., Kong, L., Xu, Y., Ge, Z., Zhao, L., Sun, J., Peng, Y., et al. (2024). General OCR theory: Towards OCR-2.0 via a unified end-to-end model. arXiv preprint arXiv:2409.01704.
Wu, Z., Chen, X., Pan, Z., Liu, X., Liu, W., Dai, D., Gao, H., Ma, Y., Wu, C., Wang, B., et al. (2024). DeepSeek-VL2: Mixture-of-experts vision-language models for advanced multimodal understanding. arXiv preprint arXiv:2412.10302.
Yin, S., Fu, C., Zhao, S., Li, K., Sun, X., Xu, T., and Chen, E. (2024). A survey on multimodal large language models. National Science Review, 11(12):nwae403.
Zapata-Sierra, A. J., Salmerón-Manzano, E., Alcayde, A., Zapata-Castillo, M. L., and Manzano-Agugliaro, F. (2023). The scientific landscape of smart water meters: A comprehensive review. Water, 16(1):113.
Zhao, W. X., Zhou, K., Li, J., Tang, T., Wang, X., Hou, Y., Min, Y., Zhang, B., Zhang, J., Dong, Z., et al. (2023). A survey of large language models. arXiv preprint arXiv:2303.18223, 1(2):1–124.
Baek, J., Kim, G., Lee, J., Park, S., Han, D., Yun, S., Oh, S. J., and Lee, H. (2019). What is wrong with scene text recognition model comparisons? dataset and model analysis. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 4715–4723.
Bai, S., Chen, K., Liu, X., Wang, J., Ge, W., Song, S., Dang, K., Wang, P., Wang, S., Tang, J., Zhong, H., Zhu, Y., Yang, M., Li, Z., Wan, J., Wang, P., Ding, W., Fu, Z., Xu, Y., Ye, J., Zhang, X., Xie, T., Cheng, Z., Zhang, H., Yang, Z., Xu, H., and Lin, J. (2025). Qwen2.5-VL technical report.
Bautista, D. and Atienza, R. (2022). Scene text recognition with permuted autoregressive sequence models. In European Conference on Computer Vision, pages 178–196. Springer.
Bochkovskiy, A., Wang, C.-Y., and Liao, H.-Y. M. (2020). YOLOv4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934.
Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., et al. (2022). On the opportunities and risks of foundation models. arXiv preprint arXiv:2108.07258.
Calefati, A., Gallo, I., and Nawaz, S. (2019). Reading meter numbers in the wild. In 2019 Digital Image Computing: Techniques and Applications (DICTA), pages 1–6. IEEE.
Carvalho, R., Melo, J., Graça, R., Santos, G., and Vasconcelos, M. J. M. (2023). Deep learning-powered system for real-time digital meter reading on edge devices. Applied Sciences, 13(4):2315.
Gómez, L., Rusinol, M., and Karatzas, D. (2018). Cutting sayre’s knot: reading scene text without segmentation. application to utility meters. In 2018 13th IAPR International Workshop on Document Analysis Systems (DAS), pages 97–102. IEEE.
Gu, Z., Zhu, B., Zhu, G., Chen, Y., Tang, M., and Wang, J. (2024). AnomalyGPT: Detecting industrial anomalies using large vision-language models. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 1932–1940.
JaidedAI (2020). EasyOCR: Ready-to-use OCR. [link].
Laroca, R., Araujo, A. B., Zanlorensi, L. A., De Almeida, E. C., and Menotti, D. (2021). Towards image-based automatic meter reading in unconstrained scenarios: A robust and efficient approach. IEEE Access, 9:67569–67584.
Laroca, R., Barroso, V., Diniz, M. A., Gonçalves, G. R., Schwartz, W. R., and Menotti, D. (2019). Convolutional neural networks for automatic meter reading. Journal of Electronic Imaging, 28(1):013023–013023.
Levenshtein, V. I. et al. (1966). Binary codes capable of correcting deletions, insertions, and reversals. In Soviet Physics Doklady, volume 10, pages 707–710. Soviet Union.
Liao, S., Zhou, P., Wang, L., and Su, S. (2019). Reading digital numbers of water meter with deep learning based object detector. In Chinese Conference on Pattern Recognition and Computer Vision (PRCV), pages 38–49. Springer.
Morales, F. (2020). Keras-OCR: A packaged and flexible version of the CRAFT text detector and keras CRNN recognition model. [link].
Mueez, A., Baranwal, A., Chaj-Mejia, J., Bhatia, G., Voelker, J. T., and Vyas, S. (2026). Vision-language models for analog gauge reading: An empirical study of specialization, transfer and reliability. arXiv preprint arXiv:2608.17723.
Rodríguez, M., Berdugo, G., Jabba, D., Calle, M., and Jimeno, M. (2014). Hd-mr: A new algorithm for number recognition in electrical meters. Turkish Journal of Electrical Engineering and Computer Sciences, 22(1):87–96.
Shi, B., Bai, X., and Yao, C. (2016). An end-to-end trainable neural network for image-based sequence recognition and its application to scene text recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(11):2298–2304.
Shu, D., Ma, S., and Jing, C. (2007). Study of the automatic reading of watt meter based on image processing technology. In 2007 2nd IEEE Conference on Industrial Electronics and Applications, pages 2214–2217.
Smith, R. (2007). An overview of the tesseract OCR engine. In Ninth International Conference on Document Analysis and Recognition (ICDAR 2007), volume 2, pages 629–633. IEEE.
Spichkova, M., Van Zyl, J., Sachdev, S., Bhardwaj, A., and Desai, N. (2019). Easy mobile meter reading for non-smart meters: Comparison of aws rekognition and google cloud vision approaches. arXiv preprint arXiv:1910.12617.
Tong, R., Liu, J., Wang, T., Hu, X., Liu, S., Wang, L., and Xu, J. (2026). Does bigger mean better? comparative analysis of CNNs and biomedical vision-language models in medical diagnosis. In 2026 International Conference on Artificial Intelligence, Computer, Data Sciences and Applications (ACDSA), pages 1–6.
Vanetti, M., Gallo, I., and Nodari, A. (2013). Gas meter reading from real world images using a multi-net system. Pattern Recognition Letters, 34(5):519–526.
Wang, F., Weng, C., Wang, X., Chen, Z., Zhao, Z., and Tang, J. (2025). Dialbench: Towards accurate reading recognition of pointer meter using large foundation models. arXiv preprint arXiv:2511.21982.
Wang, Y., Chen, Q., Hong, T., and Kang, C. (2019). Review of smart meter data analytics: Applications, methodologies, and challenges. IEEE Transactions on Smart Grid, 10(3):3125–3148.
Wei, H., Liu, C., Chen, J., Wang, J., Kong, L., Xu, Y., Ge, Z., Zhao, L., Sun, J., Peng, Y., et al. (2024). General OCR theory: Towards OCR-2.0 via a unified end-to-end model. arXiv preprint arXiv:2409.01704.
Wu, Z., Chen, X., Pan, Z., Liu, X., Liu, W., Dai, D., Gao, H., Ma, Y., Wu, C., Wang, B., et al. (2024). DeepSeek-VL2: Mixture-of-experts vision-language models for advanced multimodal understanding. arXiv preprint arXiv:2412.10302.
Yin, S., Fu, C., Zhao, S., Li, K., Sun, X., Xu, T., and Chen, E. (2024). A survey on multimodal large language models. National Science Review, 11(12):nwae403.
Zapata-Sierra, A. J., Salmerón-Manzano, E., Alcayde, A., Zapata-Castillo, M. L., and Manzano-Agugliaro, F. (2023). The scientific landscape of smart water meters: A comprehensive review. Water, 16(1):113.
Zhao, W. X., Zhou, K., Li, J., Tang, T., Wang, X., Hou, Y., Min, Y., Zhang, B., Zhang, J., Dong, Z., et al. (2023). A survey of large language models. arXiv preprint arXiv:2303.18223, 1(2):1–124.
Publicado
19/10/2026
Como Citar
TOZEVICH, Luís Gustavo Werle; SILVA, Luan Portella da; LUNARDI, Gabriel Machado; SILVEIRA, Thiago Lopes Trugillo da; OLIVEIRA, Adriano Quilião de.
Zero-Shot Recognition of Utility Meter Readings Using Vision-Language Models and General-Purpose OCRs. In: ENCONTRO NACIONAL DE INTELIGÊNCIA ARTIFICIAL E COMPUTACIONAL (ENIAC), 23. , 2026, Cuiabá/MT.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 492-502.
ISSN 2763-9061.
DOI: https://doi.org/10.5753/eniac.2026.30034.
