Unlock instant, AI-driven research and patent intelligence for your innovation.

Virtual food box defect generation method and system based on neural network

A neural network and food box technology, applied in the field of defect image sample generation, can solve problems such as the inability to meet the diversity and randomness of defects in model samples, the inability to meet the large demand for defect model samples, and the low efficiency of manual modeling, so as to avoid Decreased model fidelity, randomness, realistic effect of defect model

Pending Publication Date: 2022-04-15
GUANGDONG UNIV OF TECH +1
View PDF9 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] At present, the workload of manual modeling is huge, and the types of defects drawn are single, which cannot meet the diversity and randomness of defects in the model samples. At the same time, the efficiency of manual modeling is low, especially for the drawing of irregular defects. The effect is not good, and it cannot meet the large demand for defective model samples

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
  • Virtual food box defect generation method and system based on neural network
  • Virtual food box defect generation method and system based on neural network
  • Virtual food box defect generation method and system based on neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach 1

[0067] See figure 1 , image 3 Distance figure 1 A flow chart of a method of generating a virtual food box defective method based on a neural network according to an embodiment of the present application, image 3 A method of generating a virtual food cassette defect generating method based on a neural network according to an embodiment of the present application embodiment. The generating method is applied to the server, including the following steps:

[0068] S1, the image data is acquired by the generating counterfeit network model obtained by pre-training;

[0069] In step S1, the counter network model is generated, which is obtained by improved the existing counter network model. Its improvement is mainly: can train the generator and discrimeria in the network by filling the surface defect image data set of the real food box, and the generator and discriminator in the network are trained to generate a counterfeit network model.

[0070] The method of obtaining the defective ima...

Embodiment approach 2

[0116] See figure 2 Distance figure 2 A structural diagram of a virtual food box defective system based on a neural network is provided in the present application embodiment, including:

[0117] The first acquisition module 10 is used to obtain image data by a pre-training resulting counterfeit network;

[0118] Generating a counterfeit network model is obtained by improving the existing counterfeit network model. Its improvement is mainly: can train the generator and discrimeria in the network by filling the surface defect image data set of the real food box, and the generator and discriminator in the network are trained to generate a counterfeit network model.

[0119] The method of obtaining the defective image data set is: the RGB color image of the real food cassette defective image size is 416 * 416, and the number of sheets can be selected according to the actual situation, and the present application is preferably 180 sheets, and there is a total of 6 Class known defect ty...

Embodiment approach 3

[0151] Figure 4 A structural block diagram of an electronic device according to an embodiment of the present application is shown.

[0152] The aforementioned embodiment describes a method and system based on a virtual food box defective generating method and a system of neural networks. In one possible design, the aforementioned virtual food box defective method and system based on neural networks can be integrated into electronic devices. Such as Figure 4 As shown in this electronic device 500, the electronic device 500 can include a processor 501 and a memory 502.

[0153]The memory 502 is configured to store a procedure for performing data processing methods or resource allocation methods in any of the above embodiments, the processor 501 being configured to perform programs stored in the memory 502.

[0154] The memory 502 is for storing one or more computer instructions, wherein the one or more computer instructions are performed by the processor 501 to achieve the followin...

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 defect image sample generation, in particular to a virtual food box defect generation method and system based on a neural network. The method comprises the steps of obtaining image data through a generative adversarial network obtained through pre-training; screening the image data to obtain a first sample and a second sample; obtaining an original map; wherein the original chartlet is obtained by receiving a simulation defect image output by a generative adversarial network through an image segmentation network model and extracting features; and based on the original map, converting the original map into a normal map, and outputting the normal map to a pre-selected area in the three-dimensional virtual food box to realize defect generation of the virtual food box. The image data is generated through the generative adversarial network, then the first sample and the second sample are screened out to train the image segmentation network, the generative adversarial network and the image segmentation network model can output a large number of simulation defect images, a large number of food box defect data sets are generated, and the problem that the defect model sample demand is insufficient is solved.

Description

Technical field [0001] The present application relates to the field of defective image sample generation technology, and in particular, there is a method and system for virtual food box defects based on neural network. Background technique [0002] Defect images are often used in industrial simulations. Industrial simulation is a virtual to the entity industry. Transform the various modules in the entity industry into data into a virtual system, simulate every work and process in industrial operations in this system, and Implement various interactions. [0003] In recent years, with the rapid development of industrial materials, the factuality of models in industrial simulation technology is getting higher and higher. However, there are still many problems with high-precision with traditional methods. E.g: [0004] At present, hand-modeling workload is huge, and the defect types are drawn, and the diversity and randomness of defects in model samples cannot be met. The effect is n...

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
IPC IPC(8): G06T7/00G06T7/11G06T15/04G06N3/04G06N3/08G06V10/74G06V10/774G06V10/764G06V10/82G06K9/62
Inventor 李晋芳何明桐苏健聪郑泽胜李博
Owner GUANGDONG UNIV OF TECH