Abstract
Along with the weapon research development,more and more weaponry with high speed and high mobility are equipped in our country.In order to make target tracker of the weapons have good robustness under the situation of the target scale changed and high speed,a real-time target tracking system based on DSP+FPGA was designed.The MeanShift Algorithm was improved by using multilevel pyramid and the background weighted kernel histogram,which can reduce the influence of background in the tracking box,then the whole algorithm was optimized to implement on the embedded system in real time.The experimental results show that the design of the real-time target tracking system is robust for the rapid motion of the target,scale changing,and some degree of fuzzy.The average processing time per frame was 13.5ms,which can meet the engineering requirements of realtime and stability.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
