Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A building recognition model building method, building recognition method and device

A technology for identifying models and establishing methods, applied in the field of information processing, can solve problems such as slow algorithm speed, achieve the effect of reducing difficulty, reducing computational complexity and calculation time, and making judgments simple and fast

Active Publication Date: 2022-02-18
TUNNEL TANG TECH
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The deep learning method is currently applied in the field of data recognition, such as face recognition. In the case of a relatively large amount of data, the deep learning method has a better accuracy rate, but it is slower than the traditional algorithm. At the same time, it is currently used in construction The field of identification is still blank

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A building recognition model building method, building recognition method and device
  • A building recognition model building method, building recognition method and device
  • A building recognition model building method, building recognition method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0050] A method for establishing a building recognition model, comprising:

[0051] Use the APN data samples to train the convolutional neural network;

[0052] The convolutional neural network that has been trained is used as a building recognition model:

[0053] figure 1 An example flow of the training process of the convolutional neural network by APN data samples is shown. In various implementations of this example process, steps may be deleted, combined, or divided into sub-steps. The example process may include a preparation phase and a training phase.

[0054]In the preparation stage, it is necessary to prepare sample data for training, which includes a large amount of multi-channel data, such as tens of thousands of multi-channel image samples, and mark the correct recognition result corresponding to each sample. In this embodiment, the APN data samples include training building query formulas, positive sample building pictures, and other building pictures;

[00...

Embodiment 2

[0068] A method for building identification, comprising:

[0069] Obtain the architectural query type entered by the user;

[0070] Building feature vector acquisition;

[0071] Building identification judgment.

[0072] figure 2 An example flow of building recognition through a trained convolutional neural network is shown by inputting a building query. In various implementations of this example process, steps may be deleted, combined, or divided into sub-steps.

[0073] Specifically, the building recognition method includes the following steps 210-250:

[0074] In step 210, the building query formula input by the user is obtained; specifically, the building query formula is a building picture, and may also be a description of the picture or a certain part of the overall building picture.

[0075] In steps 220-240, input the building query formula into the training model, extract and transform it into a building feature vector of the building query formula through the b...

Embodiment 3

[0079] Such as image 3 As shown, a training device for applying a building recognition model, including:

[0080] A sample training library, the sample training library includes APN data samples, including training building query formulas, positive sample building pictures and other building pictures;

[0081] A training unit, configured to train a convolutional neural network using the sample training library;

[0082] The model generation unit generates a building recognition model through the convolutional neural network after training.

[0083] Specifically, the APN data sample is obtained through a crawler program or manual marking, or a crawler program supplemented with manual marking.

[0084] Specifically, the training unit includes:

[0085] The classification extraction unit is used to extract general features using the building recognition model, and adopts network shallow feature extraction;

[0086] A computing unit, configured to convert the general feature ...

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

No PUM Login to View More

Abstract

The invention belongs to the technical field of information processing, and relates to a method for establishing a building recognition model, a building recognition method and a device, in particular to a building recognition method based on a convolutional neural network. The present invention selects suitable APN training samples to train the convolutional neural network, transforms deep or / and shallow network features into architectural features, and forms a building recognition model; the user inputs building query formulas into the model, and can obtain building query formulas The distance between the building feature vector and the building feature vector of the image in the query database is judged and recognized, and the result is fed back to the user. Different from the existing picture recognition technology, the present invention extracts features in a specific field for buildings, reduces the difficulty of adjusting related parameters, and forms a perfect network based on convolutional neural training for building recognition. And the HNSW algorithm is used to optimize the building feature vector, which reduces the computational complexity and calculation time and makes the building recognition judgment simple and fast.

Description

technical field [0001] The invention belongs to the technical field of information processing, and relates to a method for establishing a building recognition model, a building recognition method and a device, in particular to a building recognition method based on a convolutional neural network. Background technique [0002] As one of the important artificial features, buildings are closely related to the lives of the public. The extraction and identification of their information will greatly promote the development of digital maps, building information databases, digital city modeling, virtual cities, and tourism. [0003] At present, the traditional way of extracting and identifying building information is to compare the distribution of pixels and find some possible similarities through the variance approximation of pixels. This method has low accuracy. The deep learning method is currently applied in the field of data recognition, such as face recognition. In the case of...

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
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06V20/10G06V10/764G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06V20/00G06F18/241
Inventor 张森张建辉余松林王俊明谢国兵粟彬高松贺黄学文李锦勇钟志柯许斌叶兴龙杨伟栋赫永真李征杨旭蔡贵军徐川许明敏李亚峰缪谨畅敏于长虹谭卓李星良朱子豪刘海军
Owner TUNNEL TANG TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com