基于图像处理技术的智能化人体行为识别模型研究  

Research on Intelligent Human Behavior Recognition Model Based on Image Processing Technology

作　　者：冉宪宇 RAN Xianyu(College of Computer and Software,Shaanxi National Defense Industry Vocational and Technical College,Xi’an 710300,China)

机构地区：[1]陕西国防工业职业技术学院,计算机与软件学院,陕西西安710300

出　　处：《微型电脑应用》2022年第10期175-178,共4页Microcomputer Applications

摘　　要：为了排除噪声数据影响、提高人体行为识别的准确率,从分析图像处理技术入手,对图像进行预处理之后利用混合高斯模型提取背景图像,采用改进的帧间差分算法检测目标,利用改进的k-均值算法提取图像特征,基于已获得的图像数据采用改进的HSMM模型实现人体行为识别。通过KTH人体行为数据库的实例数据验证所提出的构建模型的平均识别率高达97.5%,具有良好的检测效果,为人体行为识别提供了智能化检测手段。In order to eliminate the noise data,improve human behavior recognition accuracy,from the analysis of image processing technology,after preprocessing to the image,the mixture Gaussian model is used to extract the background image,the improved interframe difference algorithm is used to detect target,and the k-means algorithm is used to extract the image features.Based on the acquired image data with the improved HSMM model,the recognition of human behavior is realized.Through the example data of KTH human behavior database,the average recognition rate of the proposed model is as high as 97.5%,which has a good detection effect and provides an intelligent detection method for human behavior recognition.

关 键 词：图像处理 行为识别 帧间差分算法 K-均值算法 HMM模型 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]