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

Data acquisition method for AI commodity recognition training

A technology of data collection and data collection, which is applied in the field of AI commodity identification, can solve the problems of labor and energy consumption, complex data collection process, and low collection efficiency, and achieve the effect of reducing labor costs, reducing costs, and accurate results

Inactive Publication Date: 2021-01-22
上海爱购智能科技有限公司
View PDF7 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is: aiming at the deficiencies in the prior art, the present invention provides a data collection method for AI commodity recognition training, to solve the complex process of collecting data for AI commodity recognition training in the prior art, requiring a large amount of manual labor and Energy, high cost and low collection efficiency

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
  • Data acquisition method for AI commodity recognition training
  • Data acquisition method for AI commodity recognition training

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0055] This embodiment provides a data collection method for AI commodity recognition training, including the following steps:

[0056] Step 1: Establish the shopping scene database, commodity model database and model person database respectively, so that the corresponding data can be directly extracted from the database for combination according to any training requirements;

[0057] Step 1.1: Construct or collect 3D shopping scenes to establish a shopping scene database;

[0058] Step 1.1.1: According to training requirements, construct indoor scenes in professional modeling software based on real shopping scenes and / or lighting environments;

[0059] In this embodiment, if there is a clear scene requirement, first construct the indoor scene according to the real shopping scene, then import the indoor scene into the Unity engine, and simulate the lighting environment to the indoor scene in the Unity engine; if there is no clear scene requirement, you can directly Simulate the...

Embodiment 2

[0083] On the basis of the above-described embodiments, a data collection method for AI commodity recognition training is provided. In order to further better implement the present invention, the method also includes the following steps:

[0084] Step 5: Use OpenCV technology to label the annotation information of each commodity in all field of view images of the data set, and perform text supplementation, wherein the annotation information of the commodity includes at least one of the outline of the commodity and the circumscribed rectangle;

[0085] In this embodiment, labeling is text supplementary information for the generated data set, which belongs to a part of the data set, and it can be decided whether to label product information according to training requirements;

[0086] Step 5.1: Read each view image of the data set one by one;

[0087] Step 5.2: Perform threshold processing on the read field of view image;

[0088] Step 5.3: Extract the contours of the model per...

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 discloses a data acquisition method for AI commodity recognition training, which relates to the technical field of AI commodity recognition, and comprises the following steps: establishing a shopping scene database, a commodity model database and a model person database; acquiring a camera view background, combining the commodity models retrieved from the commodity model database toobtain at least one combination, and acquiring an image of each combination in the camera view background by a virtual camera to obtain a commodity data set; placing commodities in any 3D shopping scene of the shopping scene database, and setting at least one virtual camera to form a virtual shopping scene; controlling a plurality of model persons in the model person database to simulate shoppingbehaviors in the virtual shopping scene, capturing a visual field image through a virtual camera, and generating a data set; the problems that in the prior art, the AI commodity recognition training data collection process is complex, a large amount of manual labor and energy need to be consumed, the cost is high, and the collection efficiency is low are solved.

Description

technical field [0001] The present invention relates to the technical field of AI product recognition, in particular to a data collection method for AI product recognition training. Background technique [0002] With the great development of electronic payment technology, identification technology and product identification technology, more and more unmanned vending machines, unmanned retail stores, and unmanned supermarkets have appeared in the retail industry. Among them, the use of AI technology to carry out deep learning and training on products, and the AI ​​product recognition technology that accurately identifies each product in the picture and its related information is one of the key technologies for unmanned vending machines, unmanned retail stores, and unmanned supermarkets. The main trends in current development. [0003] Currently, there are two main training methods for AI product recognition. One is to place various products at different angles or directions ...

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 Applications(China)
IPC IPC(8): G06Q30/06G06Q10/06G06K9/62G06T19/00G06T13/40G06F16/2455G06F16/21
CPCG06Q30/0643G06Q10/067G06T19/006G06T13/40G06F16/21G06F16/2455G06F18/214Y02D10/00
Inventor 陈星睿商志慧
Owner 上海爱购智能科技有限公司
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