Fig 1.
The Flow Diagram of the Process of Recognition Method and Mechanism.
Fig 2.
The Photograph of RGB Behavior Video Background Subtraction.
Fig 3.
Behavior Motion Outline Model History Images of Depth Image (left) and RGB Image (right).
Fig 4.
Behavior Motion Outline Model History Images of Depth Image (left) and RGB Image (right).
Fig 5.
Category of Special Edge Pixel Circumstances.
Fig 6.
Result of Marginalized Processed Special Edge Pixels.
Fig 7.
Contrast Photographs of Some images before (left) and after (right) Marginalization.
Fig 8.
History Behavior Outline Image of Blended Depth and RGB Images.
Fig 9.
History behavior outline image integrated with depth and RGB images after 4 × 4 division.
Fig 10.
Gradient Vectors of Adjacent Edge Pixels.
Table 1.
Confusion matrix of human behavior recognition of K-means.
Table 2.
Human behavior recognition of different classification algorithms.
Fig 11.
Recognition accuracies of four algorithms experiment on DHA behavior datasets.
Fig 12.
Recognition accuracies of three algorithms experiment based on K-means on Local behavior datasets.
Table 3.
Human behavior recognition of different classification algorithms.
Fig 13.
Running time of two feature algorithms.