论文部分内容阅读
采用时一空梯度法,建立一种估计图像序列中目标偏移值的相关跟踪算法。该算法采用一维相关,而且两个方向相关运算可以并行独立运算,适于硬件实现,计算速度大大高于常规相关跟踪算法,具有很强的实时性,仿真结果还表明该算法具有较高跟踪精度,且对噪声不敏感。
A temporal gradient method is used to establish a correlation tracking algorithm to estimate the target offset in the image sequence. The algorithm uses one-dimensional correlation, and the two direction-related operations can be independently operated in parallel, which is suitable for hardware implementation. The computational speed is much higher than that of the conventional correlation tracking algorithm and has strong real-time performance. The simulation results also show that the algorithm has higher tracking Accuracy, and insensitive to noise.