| Summary: | In industrial applications, continuous monitoring of sensor data is crucial to the performance of visual inspection systems. The use of visual inspection systems is very common in the control of industrial systems. The main purpose of visual inspection systems is to measure the quality of production in industrial applications and to ensure that the quality of the production line is kept under control. In this study, we propose a system that enables real-time tracking of the values of old type analogue indicators in the industry. Older analogue displays cannot send measurement information to the inspection center numerically. The values of such indicators are collected by a human being at specific time intervals. Therefore, human surveillance is required to use these
collected data in supervision. The proposed system aims to remove the need for human surveillance by providing the numerical conversion of the values of analogue indicators. This
makes it possible to control the indicator data from a single point and detect any faults in the production line immediately. The proposed system calculates the value of indicator by finding
the starting point, the ending point and the orientation of the needle in different types of analogue indicators. The system has been tested on images of different types of indicators obtained from a factory and the internet. The analogue indicator reader system, which is tested using these images, has been shown to read the correct values with a performance of 94%.
|
|---|
