Quality Checking of Impeller using Computer Vision

Submersible pumps are indispensable in fluid transportation across industries, relying heavily on efficient impellers to ensure optimal performance. However, impeller defects can develop over time, reducing efficiency, increasing energy consumption, and costly downtimes. This project addresses the critical need for reliable defect detection in submersible pump impellers, aiming to develop an automated system capable of identifying defects. The methodology involves capturing high-resolution impeller images using an industrial-grade camera setup and applying advanced image processing techniques. The core of the defect detection system lies in implementing the Canny edge detection algorithm using Python and OpenCV, enabling precise edge extraction and defect identification. Additionally, the project incorporates machine learning algorithms to enhance the system’s accuracy and adaptability to different impeller designs and defect patterns. Real-time monitoring capabilities provide maintenance personnel with timely alerts and insights into impeller conditions, facilitating proactive maintenance strategies and minimizing the risk of unexpected breakdowns. The output of the defect detection system is displayed on an LCD display, indicating whether the impeller is defective or not. The results and findings obtained have the potential to influence industry standards and best practices for impeller defect detection and maintenance strategies. In conclusion, the development of this automated defect detection system represents a significant advancement in asset management and reliability engineering for submersible pump systems, positively impacting industry standards and maintenance practices.