A virtual interaction system and method based on skin color extraction and convolutional neural network
A technology of convolutional neural network and virtual interaction, which is applied in the field of virtual interaction system based on skin color extraction and convolutional neural network, to achieve the effect of improving recognition accuracy, improving accuracy and reducing influence
Active Publication Date: 2021-09-21
SOUTH CHINA UNIV OF TECH
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
When the user enters the image area and interacts with the system, the system uses the designed algorithm to detect and identify the effective action of the user, responds corresponding to the action mode, and feeds back to the user in the projected image information, thus forming a system composed of A complete set of visual projection interactive system composed of projection system, vision system and image recognition algorithm, but in the existing visual projection interactive system, whether it is detection robustness, anti-interference, accuracy and recognition accuracy need to be improved
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0054] The virtual interaction system and method of the present invention will be described in detail below by taking bow-drawing and archery as an example.
[0055] like figure 1 Described, a kind of virtual interactive system based on skin color extraction and convolutional neural network, comprises projector 1, near-infrared laser 4, camera 2 with near-infrared filter, common camera 3 and computer, described near-infrared laser 4 Installed along the projection surface 5, two up and down, four in total, to form a near-infrared excitation surface; the ordinary camera 3 is connected to a computer, and is used to capture the picture when the user draws the bow 7 with both hands; the near-infrared The camera 2 of the optical filter is installed on the support relative to the projection surface 5 and is electrically connected to the computer for capturing the picture when the arrow 6 touches the near-infrared laser surface, and the projector 1 is installed on the projection surfa...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a virtual interactive system and method based on skin color extraction and convolutional neural network. The system includes a projector, a near-infrared laser, a camera with a near-infrared filter, a common camera and a computer. The method includes the steps of: turning on the near-infrared laser to form a near-infrared laser surface, and the projector projects the interactive interface to the near-infrared laser surface; collecting the picture when the user makes both hands movements and the picture when the object touches the near-infrared laser surface; Process and obtain the position information of the contact point of the object, and use the distribution characteristics of skin color in the Ycbcr color space to build a Gaussian model to extract the skin color area from the picture at the moment when the hands are moved; realize it in the latent area through the human hand classifier based on the convolutional neural network Determination of the position of the hands; the initial velocity of the object is obtained through the linear relationship between the distance between the hands and the initial speed of the object. The invention improves the positioning accuracy and recognition precision, and improves the user's real sensory experience.
Description
technical field [0001] The invention relates to the fields of computer vision and image recognition, in particular to a virtual interaction system and method based on skin color extraction and convolutional neural network. Background technique [0002] Due to the rapid development of modern computer vision technology and projection display technology, the interactive system based on visual projection can create an interactive mode that enables users to have a real sensory experience. The system uses a projector to project a set of computer image information in a specific area, which expresses the output of the computer at that moment. When the user enters the image area and interacts with the system, the system uses the designed algorithm to detect and identify the effective action of the user, responds corresponding to the action mode, and feeds back to the user in the projected image information, thus forming a system composed of A complete set of visual projection intera...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F3/01G06N3/04
Inventor 谢巍张伟东刘亚文李鸿斌
Owner SOUTH CHINA UNIV OF TECH