Image processing method and device and computer storage medium

An image processing and target image technology, applied in the field of image processing, can solve the problems of semantic segmentation and segmentation accuracy reduction

Active Publication Date: 2020-04-28
QINGDAO RES INST OF BEIHANG UNIV
View PDF11 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to realize real-time semantic segmentation of images using common computers, a neural network algorithm with a relatively simple structure is usually used, but this reduces the segmentation accuracy of semantic segmentation of images.

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 processing method and device and computer storage medium
  • Image processing method and device and computer storage medium
  • Image processing method and device and computer storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0062] figure 1 is a block diagram of the hardware configuration of the image processing system 100 according to the embodiment of the present invention.

[0063] Such as figure 1 As shown, the image processing system 100 includes an image acquisition device 1000 and an image processing device 2000 .

[0064] The image acquisition device 1000 is configured to acquire a 2D image to be detected, and provide the acquired 2D image to be detected to the image processing device 2000 .

[0065] The image acquisition device 1000 may be any imaging device capable of taking pictures to obtain a target image, that is, an image to be semantically segmented, such as a camera. Alternatively, means capable of acquiring an image of an object from an imaging device.

[0066] The image processing apparatus 2000 may be any electronic device, such as a PC, a notebook computer, a server, and the like.

[0067] In this example, refer to figure 1 As shown, the image processing device 2000 may i...

Embodiment 2

[0074] An image processing method provided in this embodiment, such as figure 2 As shown, including the following steps S201-S204:

[0075] S201. Extract feature map groups of different scales of the target image, wherein each feature map in the same feature map group has different scales.

[0076] In this embodiment, the aforementioned target image is an image to be semantically segmented. This image is usually an RGB image.

[0077] In one embodiment, the above S201 is implemented through the following S2011 and S2012:

[0078] S2011. Input the target image into the convolutional neural network to obtain a first feature map of the target image for each scale in different scales.

[0079] In the first embodiment, the specific implementation of the above S2011 may be as follows: input the target image into the convolutional neural network, use the convolutional neural network to perform image feature extraction on the target image, and obtain multiple feature maps with cor...

Embodiment 3

[0152] Such as Figure 6 As shown, the embodiment of the present invention also provides an image processing apparatus 6000 . The image processing device 6000 includes: an extraction module 6100 , a fusion module 6200 , and a segmentation module 6300 . in:

[0153] The extraction module 6100 is used to extract feature map groups of different scales of the target image, wherein each feature map in the same feature map group has the same scale;

[0154] The fusion module 6200 is configured to perform feature fusion on the same scale for all the feature map groups to obtain the corresponding first fusion feature map;

[0155] The fusion module 6200 is further configured to enlarge each of the first fusion feature maps to the largest scale among the different scales, and perform feature fusion at the same scale on all the enlarged first fusion feature maps to obtain the second fusion feature map;

[0156] The segmentation module 6300 is configured to obtain a semantic segmenta...

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 an image processing method and device and a computer storage medium, and the method comprises the steps: extracting feature map groups of different scales of a target image, wherein all feature maps in the same feature map group have the same scale; performing same-scale feature fusion on all the feature map groups to obtain corresponding first fusion feature maps; amplifying each first fusion feature map to the maximum scale in different scales, and performing same-scale feature fusion on all the amplified first fusion feature maps to obtain a second fusion feature map; and obtaining a semantic segmentation result of the target image according to the second fusion feature map.

Description

technical field [0001] The present invention relates to the technical field of image processing, and more specifically, to an image processing method, an image processing device, and a computer storage medium. Background technique [0002] The semantic segmentation of images has become one of the three major tasks of computer vision. Its goal is to classify each pixel in the image, and finally divide the image into regions with different semantic meanings. Based on this, the semantic segmentation of images has a wide range of applications in medical image analysis, automatic driving, scene understanding and other fields. [0003] At present, the semantic segmentation of images is usually realized by using complex neural network algorithms. Due to the large amount of calculation of the neural network algorithm with complex structure, real-time semantic segmentation of images can only be realized by using a computer with powerful GPU capabilities to run the neural network alg...

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): G06K9/34G06K9/46G06K9/62G06N3/04
CPCG06V10/267G06V10/464G06N3/045G06F18/253
Inventor 梁晓辉卢杨于洋王平平冷芝莹
Owner QINGDAO RES INST OF BEIHANG UNIV
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