results are shown in Table 2. It can be seen that the accuracy rate based on projection
method and BP neural network method is 90.1%, and the accuracy rate based on Canny
edge detection, Hough transform, and template matching method is 93.3%, and the
accuracy of the proposed method is 98.0%. There are obvious advantages in the
proposed method.
4 Conclusion
This paper proposes a method based on convolutional neural network for water meter
reading area detection, which lays a foundation for the recognition of readings. In this
paper, the convolutional neural network is used to predict the reading area, and then the
NMS algorithm is used to remove highly overlapping results from prediction region
results to obtain the position of the reading area. The experimental results show that the
IoU of the images of 1000 test sets are all above 0.8 and combined with the three-layer
BP neural network for character recognition, the accuracy rate reaches 98.0%. The
method of this paper achieves the detection of water meter readings efficiently and
accurately. The method is not only unsusceptible to external environmental factors, but
also has generalization ability, and the model structure is simple, requiring less computing resources.
Fig. 3. Predicted results of readings
Table 2. Distribution of IoU
Method
Accuracy rate
(%)
Method based on projection and BP neural network
90.1
Method based on Canny edge detection, Hough transform, and template
matching
93.3
Proposed method with BP neural network
98.0
178
J. Wu et al.