【摘要】 基于目标颜色特征,将遗传算法和粒子滤波器相结合进行非刚性目标的实时跟踪:一般情况下,采用遗传算法跟踪目标,以最优个体作为目标状态;当发生较严重遮挡时,最优个体不一定是目标的真实状态,利用粒子滤波器的思想,以各个体的加权平均作为跟踪结果来克服遮挡影响.实验结果表明该混合算法具有较强的鲁棒性,能有效实现复杂场景下的目标跟踪.更多还原

【Abstract】 The genetic algorithm is combined with particle filter to track non-rigid target real-time based on the color feature.Generally,the target is tracked by genetic algorithm,using the optimal chromosome as the target state.When the heavy occlusion occurs,the optimal chromosome is not necessarily the real target state.Then the idea of particle filter is adopted by using the weighted mean of all the chromosomes as the tracking result to overcome the influence of occlusion.Experimental results indicate that the proposed method is robust and has good performances under complicated background.更多还原