1. The study of instrument recognition based on convolutional neural network
- Author
-
Su-Li Yan, Zhe Gao, Ze-Tian Wang, Zhao-Xue Wu, Le Hua, and Dan Wu
- Subjects
Power transmission ,Transformation (function) ,Artificial neural network ,Computer science ,business.industry ,Deep learning ,Normalization (image processing) ,Image processing ,Pattern recognition ,Artificial intelligence ,business ,Convolutional neural network ,Field (computer science) - Abstract
At present, the digital meter recognition is widely used in the field of power transmission and transformation, but rarely used in the petrochemical industry. This paper combines traditional image processing techniques with deep learning methods and proposes a meter recognition method based on an improved neural network model. By establishing a neural network model architecture, setting up the three convolutional layers and adding a batch normalization processing operation, the digital meters of the oil extraction platform are recognized, and the recognition rate is about 99.17%, which achieve the expectations. Through comparative research, the model constructed in this paper has a higher recognition accuracy and reflects a better recognition effect.
- Published
- 2021
- Full Text
- View/download PDF