An Image Processing-based Motion Tracking System for Real-time Velocity Evaluation in Microfluidic Flow Monitoring

• Main Authors: Ming Ta Tu, 塗明達
• Other Authors: S. W. Chen
• Format: Others
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/55685888956281857149

碩士 === 長庚大學 === 電子工程學系 === 101 === In this paper, we presented a Real-time motion tracking system which is coded by using C++, OpenCV and Windows API to be evaluated in microfluidic flow monitoring. In general, there are no applications with exclusive velocity evaluation for microfluidic flow. For this point of view, this paper is focus on developing a theory with less complexity and high resolution measurement. The main algorithm for motion tracking system in this paper is separating into few parts. Firstly, do the edge detection for finding the contour of each microfluidic flow ball, then using those groups of contour parameters to fit into ellipse for finding the position of balls and direction. After we got the position of balls, then we can start to track the balls and analyze the velocity of balls. For verifying the accuracy of monitoring result, we design a microfluidic flow simulated image in which we knew the result already, and use it to determine whether the result can match with the design data or not. Finally, we chose a water-in-oil microfluidic image to detect the velocity of microfluidics and to prove that whether the system can work correctly or not. The result shows that the system can track the velocity successfully. The system can trace the velocity for a single or multiple objects in images.