Detection and tracking of sphere markers

摘要: Tracking and detection of sphere markers captured by a GoPro Hero 4 Silver camera at 240 frames per second is studied discussed in this report. Different tools are designed such as, optimized color detection, edge circle point tracking. A number have been used as Kernelized Correlation Filters, Optical Flow Circle Hough Transform. The major problem KCF trackers their incapability adjusting the scale changes target. It solved designing tracker unit. marker for study spherical gray. Color detector algorithms normally reliable specific colors. In worst case they cannot detect gray objects or it comes with high error. learning algorithm was to optimize range HSV signature marker. Transform well-known method detecting circles. This function accepts many inputs which effect position circle. developed its minimizing defined error function. Finally, results smoothen applying Kalman Filter, an extremely accurate filter control industry robotics smooth predict robots. work we discuss robustness against brightness that shows how would perform more real-world conditions. final compared accuracy versus three other algorithms: CSRT, Boosting, Median Flow. comparison proposed fact fewer failures.