当箱粒子滤波算法在噪声环境下对目标进行检测与跟踪时,由于量测噪声分布不合理,导致广义似然函数表达精度不高,提出了一种改进的箱粒子滤波目标跟踪算法。该算法以高斯分布表示区间噪声,从箱粒子滤波的预测与更新步骤出发,在高斯分布环境下修改广义似然函数,推导了伯努利箱粒子滤波更新过程的表达式。
When box-particle filter algorithm is used for target detection and tracking in the noisy environment,the expression of the generalized likelihood function is not accurate due to the unreasonable measurement noise distribution.So an improved target tracking algorithm based on box-particle filter is proposed.In this algorithm,the interval noise is represented by Gaussian distribution.Starting from the prediction and update step,the expression of the Bernoulli box-particle filter in Gaussian distribution is derived by modifying the generalized likelihood function.