International Journal Of Imaging

Abstract

Automated Detection of oil and water interfaces using Deep Learning- based sensor signal classification

Author: Nalla Sanjana Reddy, Kurma Bunny , Gopalapuram Abinav , K. Santhi

Abstract:

The detection of oil and water interfaces is a critical requirement across various industries, including petroleum production, chemical processing, and environmental monitoring. Accurate identification of these interfaces is essential for process optimization, safety, and environmental compliance. Traditionally, the task has relied on manual observations or basic sensor-based systems, such as capacitance, resistive, or ultrasonic sensors, to detect changes in material properties. While effective in controlled environments, these traditional methods face significant challenges in dynamic, noisy, and complex industrial settings, where factors like emulsions, temperature variations, and pressure fluctuations can affect sensor performance. The limitations of traditional systems highlight the need for more robust and intelligent solutions. Recent advances in artificial intelligence, particularly deep learning, have opened new opportunities to enhance the performance of sensor-based sy