Computer Vision Based Energy Monitoring System using Meter Image Capturing System (MICAPS)

In the era of smart grid, extraction of data from meters is the primary step. Conventionally data extraction from non-communicable meters are done manually. In order to provide complete smart solutions to the customers the utilities are planning for a mass roll out smart meters which is costly. Approximately |3000 to |4000 is required for replacing a noncommunicable meter to communicable meter and also requires to hamper the existing infrastructure. In this context this paper proposes a meter image capturing system (MICAPS). MICAPS is a image based technique for extraction of data. MICAPS is formed by integrating raspberry pi with raspberry pi camera. Python code is used to configure the camera and take the images and send it to cloud. This image is a seven segment display. In order to convert a seven segment display to text a computer vision optical recognition algorithm is used. The overall cost of the MICAPS is |4000 with which it is possible to monitor minimum 2 meters. One more advantage with MICAPS system is that remotely the operator will be able to change the granularity of the data based on the required. Thus this MICAPS will help in monitoring the energy in a community based distribution networks.