摘 要：引入改进的隐马尔可夫模型算法，针对真实场景中运动目标轨迹的复杂程度对各个轨迹模式类建立相应的隐马尔可夫模型，利用训练样本训练模型得到可靠的模型参数；计算测试样本对于各个模型的最大似然概率，选取最大概率值对应的轨迹模式类作为轨迹识别的结果，对两种场景中聚类后的轨迹进行训练与识别。实验结果表明，平均识别率分别达到87.76 %和94. 19%。
trajectory recognition of moving objects based on hidden markov model
pan qiming1,cheng yongmei2
（1.dept. of automation, zhongshan institute, university of electronic science & technology of china, zhongshan guangdong 528402, china;2.college of automation, northwestern polytechnical university, xi’an 710072, china）
abstract:using modified hidden markov model, firstly, aiming at the complex degree of the objects’ trajectories in real scene, the models were built for every trajectory pattern, and the training samples were used to get the credible parameters of the model.finally, the maximum likelihood probability of test samples were computed to all of the trained model, the maximum value was saved and the corresponding model was the recognition result. then train and recognize the samples clustered, and average recognition rate reach 87.76 % and 94. 19% respectively.......