<![CDATA[The use of PDAM (Regional Drinking Water Company) water meters has become the standard for measuring water consumption by customers in Indonesia. However, the process of reading water meters is still mostly done manually by officers, which can cause various problems. For example, reading errors can occur due to human factors or environmental conditions, such as poor lighting or a dirty water meter. Additionally, this process requires a lot of time and effort and has the potential to lead to fraud. To overcome this challenge, this research focuses on developing a digit detection system for PDAM water meters using the YOLOv5 deep learning model. Using a dataset covering various lighting conditions and viewing angles, the model is trained to recognize and classify the digits on water meters. Initial results show that this model can produce accurate predictions, with high levels of precision and recall. However, more testing and evaluation are needed to ensure that these systems can perform well in real-world conditions.]]>
Copyrights © 2023
