论文部分内容阅读
提出一种广义霍夫变换方法对相交目标进行轨迹预测,该方法利用概率分布原理,在参数空间中寻找点分布最集中的区域,以确定目标轨迹二次曲线方程的系数。为了提高搜索速度,提出采用投影法定位二次曲线方程的系数,从而使该方法满足实时性要求。该方法抗噪性强,克服了最小二乘轨迹预测法对噪声敏感的缺点。实验中采用最小二乘法和广义霍夫变换对相交目标进行轨迹预测,目标分离时正确匹配率分别为87.5%和95%。
A generalized Hough transform method is proposed to predict the trajectory of intersecting targets. This method uses the principle of probability distribution to find the region with the most point distribution in the parameter space to determine the coefficient of the quadratic curve equation of the target trajectory. In order to improve the search speed, the projection coefficients are put forward to locate the quadratic equation so that the method can meet the real-time requirements. The method is robust against noise and overcomes the shortcomings of the least squares locus prediction method being sensitive to noise. In this experiment, the least square method and the generalized Hough transform are used to predict the trajectory of the intersecting targets. The correct matching rates are 87.5% and 95% respectively.