利用广义霍夫变换对相交目标进行轨迹预测

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%。

The track forecast of intersectant objects by general Hough transform

A kind of general Hough transform method was proposed for track forecast of intersectant object.It makes use of the theory of probability distributing,and search the focus of point distributing in the parameter space,to get hold of the coefficient of conic equation of object track.In order to increase the speed of search,projection method was proposed to locate the coefficient of conic equation.As a result,this method can attain real time.This kind of general Hough transform method has the characteristic of antinoise character,and overcome the shortcomings of being sensitive to noise that least squares estimation possesses.In experiments,least square estimation and general Hough transform method were compared for track forecast of intersectant objects.When intersectant objects separated each other,the correct probability of matching was 87.5% and 95% respectively.