Image-based human body border recognition method
A frame recognition and image technology, applied in the field of image recognition, can solve problems such as the inability to quickly identify the frame of a character, and achieve the effect of simplified calculation and real virtual reality.
Active Publication Date: 2019-01-11
重庆虚拟实境科技有限公司
View PDF7 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] The present invention provides an image-based human body frame recognition method to solve the problem that the character frame in the character image cannot be quickly recognized in the prior art
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 Construction
[0016] In order to make the technical means, creative features, goals and functions achieved by the present invention clearer and easier to understand, the present invention will be further elaborated below in conjunction with the accompanying drawings and specific embodiments:
[0017] The present invention proposes an image-based human frame recognition method, comprising the following steps:
[0018] S1. Obtaining the collected image through the collection of the image collection device;
[0019] S2. The external image processing device receives the video stream collected by the image acquisition device, and performs matting processing on each frame of image in the video stream, so that the background in the collected image is removed and only the characters are left to obtain the matted image;
[0020] S3. According to the matting image, the external image processing device equalizes the transparent channels A of all background pixels except the characters in each frame of...
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 provides an image-based human body border recognition method, comprising the following steps: S1, an image acquisition device acquires an acquired image; S2, an external image processingdevice receives the video stream collected by the image acquisition device, and performs matting processing on each frame image in the video stream; S3, according to the matting image, the external image processing device equals the transparent channel A of all background pixels except the person in each frame image in the video stream to 0; S4, the external image processing device forms each frame image in the video stream into a transparent background person image; S5, the CPU receives the transparent background person image of the external image processing device and transmits the transparent background person image to the GPU, and the GPU calculates the person border of each transparent background person image according to the transparent channel A value of all pixels in each transparent background person image. The image-based human body border recognition method solves the problem that a human body border in a human image cannot be recognized quickly in the prior art.
Description
technical field [0001] The invention relates to image recognition technology, in particular to an image-based human frame recognition method. Background technique [0002] The Chinese patent discloses a high-accuracy multi-part human body recognition method based on convolutional neural network with the application number CN201610056618.9. The method includes: making full use of the depth information of the image, which greatly improves the accuracy of image recognition Secondly, based on the RCNN method, the Selective Search algorithm is used to form the candidate frame, which is more suitable for the accuracy of positioning information under the deep convolutional network than the sliding window method; further, by replacing the last layer of the convolutional neural network with the Softmax layer SVM finally obtains a score based on classification; in addition, after obtaining the SVM score of each candidate frame relative to each category, a candidate frame based on over...
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 IPC IPC(8): G06K9/00G06K9/46
CPCG06V20/41G06V10/44
Inventor 周鑫陶澍闫旭
Owner 重庆虚拟实境科技有限公司