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

Multi-object content joint image generation method and system

An image generation and object image technology, applied in the field of computer vision, can solve the problems of lack of organic integration and loss of creativity in the field of art, and achieve the effect of improving the speed of solving cases, improving user experience and great economic value.

Pending Publication Date: 2022-07-22
SHANDONG UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The former's fusion mechanism can only generate a few specified types of images, and relies too much on human prior knowledge. The difference between the original image and the generated image is usually subtle, resulting in the loss of unique creativity in the art field; while the latter's Fusion can reasonably distribute and arrange multiple completely unrelated objects in a picture, thus filling the gaps in human imagination
[0005] That is to say, most of the current image fusion is to extract features from the same type of images, and then fuse the respective features, but lacks the operation of organically integrating the content level of multiple irrelevant object images directly.

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
  • Multi-object content joint image generation method and system
  • Multi-object content joint image generation method and system
  • Multi-object content joint image generation method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0059] like Figure 1-Figure 5As shown, this embodiment provides a multi-object content joint image generation method, which includes the following steps:

[0060] Step S1: obtaining original images of various objects and preprocessing;

[0061] Step S2: using the pre-trained Transformer model to perform dimension compression on the preprocessed original images of various objects to obtain one-dimensional vectors of the images of the multiple objects;

[0062] Step S3: adding the one-dimensional vectors of multiple object images to obtain a multi-object image addition vector;

[0063] Step S4: Based on the multi-object image addition vector, a pre-trained generative adversarial network is used to obtain a multi-object fusion image.

[0064] In step S2, the described process of pre-training the Transformer model is specifically:

[0065] For the preprocessed original image of the object, use the encoder to obtain the one-dimensional vector of the original image of the object...

Embodiment 2

[0148] This embodiment provides a multi-object content joint image generation system, including:

[0149] an image acquisition module, configured to acquire original images of various objects and perform preprocessing;

[0150] The image dimension reduction module is configured to use the pre-trained Transformer model to perform dimension compression on the preprocessed original images of various objects to obtain one-dimensional vectors of the multiple object images;

[0151] The image vector addition module is configured to add one-dimensional vectors of multiple object images to obtain a multi-object image addition vector;

[0152] The image fusion module is configured to add vectors based on multi-object images and obtain multi-object fusion images by using a pre-trained generator network.

[0153] Examples and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the content disclosed in the first embodiment a...

Embodiment 3

[0157] This embodiment provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, implements the steps in the method for generating a multi-object content joint image as described in the above-mentioned first embodiment.

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 provides a multi-object content joint image generation method and system. The method comprises the steps that original images of various objects are acquired and preprocessed; a pre-trained Transform model is used to carry out dimension compression on the preprocessed original images of the plurality of objects, and one-dimensional vectors of the plurality of object images are obtained; adding the one-dimensional vectors of the plurality of object images to obtain a multi-object image addition vector; based on the multi-object image addition vector, utilizing a pre-trained generative adversarial network to obtain a multi-object fusion image; through a Transform model and a generative adversarial network, content fusion is carried out on multiple types of different images, and the operation of directly carrying out organic fusion on the content levels of multiple irrelevant object images is realized.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a method and system for generating a multi-object content joint image. Background technique [0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art. [0003] With the continuous development of human civilization, the high-speed dissemination and large-scale application of the Internet have made the communication between people closer. The rise of various chat software and the birth of PS tools have not only promoted the communication between people The communication between them also promotes the development of productivity, and at the same time breeds a variety of images that are adapted to different scenes. However, at this stage, if you want to generate images of unrealistic scenes on a large scale, you must use computer vision. [0004] Today, most of the e...

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): G06T11/60G06T9/00G06N3/08G06N3/04G06V10/774G06V10/764G06V10/82G06K9/62
CPCG06T11/60G06T9/002G06N3/084G06N3/047G06N3/045G06F18/24G06F18/214
Inventor 陈振学叶梦婷刘龙成贲晛烨郭庆强杜付鑫
Owner SHANDONG UNIV