查詢結果分析
相關文獻
頁籤選單縮合
題名 | Abnormal Human Activity Recognition System Based on R-Transform and Independent Component Features for Elderly Healthcare= |
---|---|
作者 | Khan, Zafar Ali; Sohn, Won; |
期刊 | Journal of The Chinese Institute of Engineers |
出版日期 | 20130600 |
卷期 | 36:4 2013.06[民102.06] |
頁次 | 頁441-451 |
分類號 | 448.6 |
語文 | eng |
關鍵詞 | Abnormal human activity recognition; Elderly healthcare; Feature extraction; R-transform; |
英文摘要 | The recognition of human activities from video sequences has been an active research area from several years. This paper presents a novel abnormal human activity recognition system utilizing R transform and Independent Component Analysis (ICA) methods. Feature extraction by the combination of R transform and ICA is rarely implemented for abnormal human activity recognition. The dataset of six abnormal activities have been produced. Features extracted from the sequence of six abnormal activities: backward fall, forward fall, chest pain, headache, vomit, and faint are trained and recognized by utilizing Hidden Markov Model (HMM). Activity recognition achieved by our system shows improved recognition results as compared to the state of the art methods: Principal Component Analysis (PCA), ICA, and PCA-ICA combination. The mean recognition rate of 90.08% is obtained by using our system model. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。