Image vectorization method and system

A vectorization and image technology, applied in the field of image processing, can solve problems such as low accuracy and non-standardized information density, and achieve high recall effect

Active Publication Date: 2021-10-15
KE COM (BEIJING) TECHNOLOGY CO LTD
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the rasterized image usually has the following disadvantages: non-standardization and low information density, etc.
If similar house types are retrieved based on the above identification results, the corresponding accuracy will be very low

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
  • Image vectorization method and system
  • Image vectorization method and system
  • Image vectorization method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] Specific embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described here are only used to illustrate and explain the present invention, and are not intended to limit the present invention.

[0041] Before introducing various embodiments of the present invention, the image vectorization system involved in the present invention will be briefly described.

[0042] Such as Image 6 As shown, the image vectorization system involved in the present invention may include: an image segmentation system 10 (for example, image 3 Deep Neural Network Segmentation (DNN) module shown), combination means 20 (e.g., image 3 Integer optimization (IP) module shown) and post-processing device 30 (for example, image 3 The vectorized post-processing (Vetor-processing) module shown). The input of the image vectorization system is a rasterized floor plan. First, the i...

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 relates to the technical field of image processing, and discloses an image segmentation model training method, an image segmentation method, an image vectorization method and a system thereof. The training method includes: using a first training sample set to train the image segmentation model based on a deep neural network; determining the recognition loss of each of a plurality of preset object samples in the first training sample set; Based on the recognition loss of each of the plurality of preset object samples and the preset weight of the recognition loss of each of the plurality of preset object samples, determine the total of the first training sample set recognition loss; and adjusting parameters of the image segmentation model according to the total recognition loss of the first training sample set. The present invention can use the deep neural network to segment and obtain multiple elements of the image, so that the vectorization of the rasterized house type map can be realized with high precision, and then the high recall of similar house types can be realized through the vectorized house type map.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a training method for an image segmentation model, an image segmentation method, an image vectorization method and a system thereof. Background technique [0002] At present, the storage and circulation of most floor plans are based on rasterized images, wherein the rasterized images mainly include CAD drawings, rendered drawings, and even hand-drawn drawings. However, the rasterized image usually has the following disadvantages: non-standardization and low information density. If it is desired to match similar house types based on the rasterized house type diagram, different recognition methods are usually used to identify (or segment) the corresponding elements in the house type diagram, and then the house types are matched according to the identified elements. Among them, the processes of different identification methods are completely different, and the accuracy 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): G06T7/12G06T7/13G06N3/04G06N3/08G06T11/20G06F16/56G06F16/583
CPCG06T7/12G06T7/13G06N3/08G06T11/206G06F16/56G06F16/583G06T2207/20164G06T2207/20081G06T2207/20084G06N3/045
Inventor 李雨龙
Owner KE COM (BEIJING) TECHNOLOGY CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap