Correlation-Based Template Tracking Of Moving Object

This paper presents the correlation-based motion estimation technique for 3D displacement of objects. Two high-speed cameras are configured as a stereovision system and synchronized in real-time. Finger hand motions captured form digital images at 1500 fps 2000 respectively. A complete acquisition is calibrated to determine intrinsic extrinsic parameters which were later used correlation algorithm. The grayscale image frames acquired from correlated using square templates 10x10 pixels created reference image. finger discussed with varying camera speed measure brightness inconsistency. observations coefficient indicate that proposed algorithm efficient up 20 50 cases was increased 0.987 0.972 left right cameras, respectively, while 0.924 0.898. developed MATLAB validated by tracing sinusoidal solid rectangular element an accelerometer sensor mounted over block.