An indoor self-positioning method and device based on discrete grid and image matching
A discrete grid and image matching technology, applied in neural learning methods, image data processing, biological neural network models, etc., can solve problems such as the inability of PlaNet system to locate, and the lack of research on image-adaptive three-dimensional position positioning, so as to achieve navigation. The full effect of technology
Active Publication Date: 2021-03-19
吉奥时空信息技术股份有限公司
View PDF8 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, the PlaNet system is designed based on a spherical discrete grid. The location information obtained by image positioning is two-dimensional, and the image set used for training is a certain geographical area, which is an outdoor scene, so that the PlaNet system cannot be used in an indoor environment. Positioning in 3D
Unfortunately, the research on image adaptive 3D position positioning for indoor environments is very scarce, so the research on image adaptive 3D position positioning for indoor environments needs to be supplemented and perfected urgently
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 approach
[0063] The specific implementation method of step S103 is as follows:
[0064] In this step, the task of image acquisition is completed mainly through manual on-site acquisition and supplemented by network image collection. On the basis of step S101 and step S102, form a grid map with the unit of 4m*4m*4m in the research area; dispatch workers to the corresponding cell position to collect indoor images, and pay attention to the interior of the room when collecting images. Distinctive objects, structures or arrangements;
[0065] Take photos at the four corners of the room toward the center of the room, so that both the global features of the room and the local unique features of the room can be obtained. The number of photos is usually kept between 4 and 10. Then the collected image data P i Add it to the corresponding record field of the table created in step S102, as the initial image data, if there is a change in the cell later, you can take a new photo to obtain the lat...
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 present invention provides an indoor self-positioning method and equipment based on discrete grid and image matching. The method includes: using an improved global discrete grid combined with a ResNet neural network method, fully considering the multi-layer and different heights of indoor buildings , divide the human activity space by global discrete grids to obtain multiple cells of different heights, and then train the prediction model specific to each cell through the neural network, and then perform matching prediction on the input picture to obtain the highest matching degree with the picture unit, so as to realize the indoor positioning of the picture; the device is used to realize the method; the beneficial effect provided by the present invention is: realizing high-precision self-adaptive three-dimensional position positioning of the indoor environment picture.
Description
technical field [0001] The invention relates to the field of image positioning, in particular to an indoor self-positioning method and equipment based on discrete grid and image matching. Background technique [0002] In the Internet age, a large amount of information data is generated every day, and these data are usually related to geographical location. If they are not organized and stored efficiently, these data cannot be retrieved and used effectively. The global discrete grid is born for this purpose. It divides the entire earth space into multiple voxels or surface units, and puts the information data into the corresponding primitives. When using it, the corresponding spatial position can be obtained by retrieving the grid. data. According to the different spatial dimensions, the global discrete grid can be divided into: spherical subdivision system and spherical subdivision system. The spherical subdivision system divides the space of the earth's surface, ignoring ...
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): G06F16/587G06T17/00G06T17/20G06F16/22G06F16/29G06N3/04G06N3/08
CPCG06F16/587G06T17/005G06T17/20G06F16/29G06F16/2282G06N3/08G06N3/045
Inventor 朱伟奇曹斌魏延峰解云飞苏俊俞鹏
Owner 吉奥时空信息技术股份有限公司