A kind of trademark image retrieval method

An image retrieval and image technology, applied in digital data information retrieval, instruments, computing, etc., can solve the problems that image features cannot express the high-level semantics of users, problems have not been solved well, and retrieval results are unsatisfactory, etc., to achieve The effect of fast calculation speed, small amount of data, and elimination of wrong matching

Active Publication Date: 2022-04-29
南昌奇眸科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Although the research on content-based trademark image retrieval is relatively active, and some systems have been put into use, there are still some problems that have not been well resolved.
The biggest difficulty is that there is no correspondence between the low-level image content features extracted by the system and the high-level semantics used by users when searching. people are satisfied
[0005] In the previous manual coding process, trademark certification personnel can weight the coding results according to certain rules. The weights corresponding to important parts of the original trademark are large, so that these important graphics in the original trademark can be highlighted, while the existing ones The system often extracts the global features of the image for matching, and cannot highlight these important information, so the reliability of the system is not high; another important problem is the speed of retrieval. As the number of images in the library increases, the speed of retrieval becomes a constraint A bottleneck of the system, and the cumulative number of existing registered trademarks in my country has exceeded 10 million

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

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0053] A trademark image retrieval method, comprising the following steps:

[0054] S1: Perform multi-scale feature extraction on the image to be retrieved and the comparison image in the image database;

[0055] S2: Carry out similarity matching between the image features to be retrieved and the compared image features between global scales;

[0056] S3: filter the correct match;

[0057] S4: candidate similar region segmentation;

[0058] S5: Similarity matching between local scales in the region, sorting the comparison images according to the size of the matching similarity.

Embodiment 2

[0060] A trademark image retrieval method, comprising the following steps:

[0061] S1: Multi-scale feature extraction is performed on the image to be retrieved and the comparison image in the image database; multiple sliding windows of different scales are used to segment the regions of the image to be retrieved and the comparison image, and image features in the sliding window window are extracted;

[0062] Include the following steps:

[0063] (1) Extract the gradient direction histogram feature of the image pixel in the sliding window window;

[0064] (2) Gradient orientation histogram quantization coding;

[0065] (3) Normalization;

[0066] (4) Spatial distribution description, and cascaded gradient direction histogram and spatial distribution.

[0067] S2: Carry out similarity matching between the features of the image to be retrieved and the features of the compared image on a global scale; slide the sliding window in the image to be retrieved, traverse all the wind...

Embodiment 3

[0074] A trademark image retrieval method, comprising the following steps:

[0075] S1: Multi-scale feature extraction is performed on the image to be retrieved and the comparison image in the image database; multiple sliding windows of different scales are used to segment the regions of the image to be retrieved and the comparison image, and image features in the sliding window window are extracted;

[0076] Include the following steps:

[0077] (1) Extract the gradient direction histogram feature of the image pixel in the sliding window window;

[0078] (2) Gradient orientation histogram quantization coding;

[0079] (3) Normalization;

[0080] (4) Spatial distribution description, and cascade gradient direction histogram and spatial distribution;

[0081] Step (1), calculate the horizontal gradient and vertical gradient of the image pixel, using the direction template [-1,0,1], the calculation rule is [G h ,G v ] = gradient(F),

[0082] Step (2), the direction angle...

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 field of image retrieval, and particularly relates to a trademark image retrieval method, comprising the following steps: performing multi-scale feature extraction on the image to be retrieved and a comparison image in an image database; performing global scale inter-relationship between the image characteristics to be retrieved and the comparison image features Similar matching; screening correct matching; candidate similar region segmentation; similar matching between local scales in the region, sorting of similar comparison images in retrieval results. The invention has high retrieval accuracy, low missed detection rate and high speed.

Description

technical field [0001] The invention belongs to the field of image retrieval, in particular to a trademark image retrieval method. Background technique [0002] Trademark plays a very important role in the industrial and commercial society. It is a symbol of a company, product or service. It is integrated with the product quality, service quality, and management of the enterprise. It becomes a symbol of corporate reputation and an intangible asset. [0003] Trademark retrieval refers to the process of finding images similar to the input trademark image from the existing trademark image database. [0004] Although the research on content-based trademark image retrieval is relatively active, and some systems have been put into use, there are still some problems that have not been well resolved. The biggest difficulty is that there is no correspondence between the low-level image content features extracted by the system and the high-level semantics used by users when searching...

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): G06F16/583
Inventor 李建圃樊晓东
Owner 南昌奇眸科技有限公司
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