摘要
因传统算法复杂且浪费大量时间,提取到的有效信息较少,所以对老年人行走异常步态特征提取算法做优化,提出积分变换下老年人行走异常步态特征提取算法。利用归一化处理对传统算法中异常步态周期做分析,发现提取有效信息少,并且计算时间长。为此,利用积分变换方法对上述算法做优化处理,先通过背景减法提取老年人异常步态信息,并利用角度射线投影变换对提取出信息做检测,使处理后算法更简便、运算时间较短,在相同效能下提取的样本信息数量多,精准率高。实验结果表明,优化后算法能迅速、高效地提取更多信息,为老年人治疗争取更多时间,避免部分死亡发生。
The traditional algorithms are complex and time consuming, and they only extract less effective information. It is necessary to optimize the feature extraction algorithm of abnormal gait for the elderly. In this article an algorithm for extracting gait features of elderly people with abnormal gait based on integral transform. After the abnormal gait period in traditional algorithm was analyzed by normalization, we found that the algorithm only extracted less effective information and needed long computing time. Therefore, we used the integral transformation method to optimize this algorithm. Firstly, we used the background subtraction to extract abnormal gait information of elderly people, and then we detected the information by angle ray projection transformation, so that the algorithm was simpler and more accurate, which required shorter operation time and extracted more sample information under the same efficiency. Simulation results prove that the optimized algorithm can quickly and efficiently extract more information, we can buy more time for the treatment of the elderly and avoid some deaths.
作者
万康
WAN Kang(Physical Education College,Zhengzhou University,Zhengzhou Henan 450000,China)
出处
《计算机仿真》
北大核心
2020年第10期429-432,460,共5页
Computer Simulation
关键词
老年人
行走异常
步态特征
特征提取算法
信息准确率
Elderly people
Walking abnormality
Gait characteristics
Feature extraction algorithm
Information accuracy