Scholarlink

<p>In this study, the capability of Event-based Vision Sensor (EVS) cameras for measuring simultaneously bubble motion, size, and shape in complex bubbly flows has been assessed. This assessment initially focused on benchmark evaluation of the accuracy of an EVS camera in tracking the motion and size of precisely manufactured naturally buoyant particles in quiescent water. Experiments were carried out in synchronization with a conventional high-speed (HS) camera, enabling quantitative comparison. The synchronized measurements obtained from both cameras agreed well with differences of less than 0.65% in measured particle velocities and 2.4% in measured diameters, demonstrating the EVS camera’s capability to precisely measure motion and dimensions. Following this validation, the assessment expanded to characterize bubbles generated by injecting compressed air at various pressures at the bottom of a quiescent water tank. The EVS camera demonstrated comparable performance to the HS camera in measuring bubble size and velocity at low to moderate injection pressures. The bubble tracks uniquely captured by the EVS camera, due to its adjustable accumulation time, can be leveraged to extract trajectory and velocity information, as well as instances of partial bubble visual overlap in the camera view. However, event saturation is observed at the EVS camera for several events generated by bubble motion that surpasses a critical threshold (e.g., dense bubbly flow). Subsequently, the increased latency at both pixel levels and read-out leads to temporal misalignment relative to the flow bubble dynamics. This challenge was addressed by employing a pulsed illumination source, which allowed the camera to capture dense bubbly flows, as the increased presence of bubbles corresponds to a reduced number of EVS recorded events. Overall, the assessment highlights that EVS cameras offer comparable performance to conventional HS cameras in characterising bubbly flows with two additional unique advantages. (i) Ability to record bubble tracks, which can be directly used for bubble velocimetry. (ii) An instantaneous EVS camera image records events at a significantly reduced data rate, which enables real-time bubble EVS imaging at higher speeds than conventional HS cameras.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Assessment of event-based vision sensor cameras for measuring bubbly flow characteristics

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月30日

查看全文
  • Ahmed Al Brahim,
  • Kuppuraj Rajamanickam,
  • A. M. K. P. Taylor,
  • Yannis Hardalupas

摘要

In this study, the capability of Event-based Vision Sensor (EVS) cameras for measuring simultaneously bubble motion, size, and shape in complex bubbly flows has been assessed. This assessment initially focused on benchmark evaluation of the accuracy of an EVS camera in tracking the motion and size of precisely manufactured naturally buoyant particles in quiescent water. Experiments were carried out in synchronization with a conventional high-speed (HS) camera, enabling quantitative comparison. The synchronized measurements obtained from both cameras agreed well with differences of less than 0.65% in measured particle velocities and 2.4% in measured diameters, demonstrating the EVS camera’s capability to precisely measure motion and dimensions. Following this validation, the assessment expanded to characterize bubbles generated by injecting compressed air at various pressures at the bottom of a quiescent water tank. The EVS camera demonstrated comparable performance to the HS camera in measuring bubble size and velocity at low to moderate injection pressures. The bubble tracks uniquely captured by the EVS camera, due to its adjustable accumulation time, can be leveraged to extract trajectory and velocity information, as well as instances of partial bubble visual overlap in the camera view. However, event saturation is observed at the EVS camera for several events generated by bubble motion that surpasses a critical threshold (e.g., dense bubbly flow). Subsequently, the increased latency at both pixel levels and read-out leads to temporal misalignment relative to the flow bubble dynamics. This challenge was addressed by employing a pulsed illumination source, which allowed the camera to capture dense bubbly flows, as the increased presence of bubbles corresponds to a reduced number of EVS recorded events. Overall, the assessment highlights that EVS cameras offer comparable performance to conventional HS cameras in characterising bubbly flows with two additional unique advantages. (i) Ability to record bubble tracks, which can be directly used for bubble velocimetry. (ii) An instantaneous EVS camera image records events at a significantly reduced data rate, which enables real-time bubble EVS imaging at higher speeds than conventional HS cameras.