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

Image region extraction result identification method and device, and computer readable storage medium

A technology of image area and identification method, which is applied in the field of mobile communication, and can solve problems such as image area affecting the normal operation of applications, user function troubles, image area extraction failure, etc.

Pending Publication Date: 2021-07-30
NUBIA TECHNOLOGY CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, at this stage, when the image extraction of a specific area is carried out through deep learning and other methods, the image area extraction failure may occur. At the same time, the failure of the image area extraction will also affect the normal operation of related applications, or bring functional use on the trouble
At this stage, there is no good way to detect the specific application of the image region extraction results.

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 region extraction result identification method and device, and computer readable storage medium
  • Image region extraction result identification method and device, and computer readable storage medium
  • Image region extraction result identification method and device, and computer readable storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0081] image 3 It is a flow chart of the first embodiment of the image region extraction result recognition method of the present invention. A method for identifying an image region extraction result, the method comprising:

[0082] S1. Acquire a first edge image and a second edge image obtained after edge extraction of the original image and its corresponding mask image respectively.

[0083] S2. For each pixel of the second edge image, determine a first number of pixels whose pixel value is not less than a preset pixel value, and at the same time, for each pixel of the second edge image, determine that the pixel value is equal to the The second number of pixels of the preset pixel value.

[0084] S3. For each pixel of the second number, determine each corresponding pixel in the first edge image, and determine that all grayscale values ​​are equal to the predetermined Set the pixel value of the third number of paired pixels.

[0085] S4. If the ratio of the third number ...

Embodiment 2

[0097] Figure 4 It is a flow chart of the second embodiment of the image region extraction result recognition method of the present invention. Based on the above embodiment, the acquisition of the first edge image and the second edge image obtained after edge extraction of the original image and its corresponding mask image respectively, include:

[0098] S11. Obtain the input original image.

[0099] S12. Extract an image area from the original image, and obtain the mask image corresponding to the original image through deep learning, wherein the image area is a sky area.

[0100] Optionally, in this embodiment, when the user extracts the image region of the selected image, the user inputs the original image of this embodiment from the gallery of the device or the network terminal, and when the application performs the function application based on the image region extraction, Get the original image before image region extraction.

[0101] Optionally, in this embodiment, ...

Embodiment 3

[0104] Figure 5 It is a flow chart of the third embodiment of the image region extraction result recognition method of the present invention. Based on the above embodiment, for each pixel of the second edge image, determine the first number of pixels whose value is not less than the preset pixel value pixel, and at the same time, for each pixel of the second edge image, determining a second number of pixels whose pixel value is equal to the preset pixel value includes:

[0105] S21. Acquire regional features of the image region.

[0106] S22. Determine the preset pixel value corresponding to the region characteristic, wherein when the image region is the sky region, the region characteristic is pure white, and the preset pixel value is 255.

[0107] Optionally, in this embodiment, the area features of the image area are acquired. For example, when the image area is the sky, the feature of this area is a white solid color feature.

[0108]Optionally, in this embodiment, the...

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 region extraction result recognition method and device and a computer readable storage medium, and the method comprises the steps: determining a first number of pixels with the pixel value not smaller than a preset pixel value for each pixel of a second edge image, and meanwhile, determining a second number of pixels of each pixel of the second edge image, determining a second number of pixels whose pixel values are equal to the preset pixel value; for each pixel of the second number, determining each corresponding pixel in the first edge image, and determining a third number of paired pixels in neighborhoods corresponding to all the corresponding pixels; and if the ratio of the third number to the first number is greater than a preset value, determining that the image region is successfully extracted. According to the invention, an efficient and accurate image area extraction result identification scheme is realized, the normal operation of related applications is prevented from being influenced by the failure of image area extraction, or the trouble in function use is prevented from being brought to a user, the productivity is improved, and the user experience is enhanced.

Description

technical field [0001] The present invention relates to the field of mobile communication, in particular to a method, device and computer-readable storage medium for identifying an image region extraction result. Background technique [0002] In the prior art, with the continuous development of intelligent terminal devices, various applications and users have higher and higher requirements for image recognition. In particular, the extraction of specific regions in an image has become the basis of many image recognition applications. However, at this stage, when the image extraction of a specific area is carried out through deep learning and other methods, the image area extraction failure may occur. At the same time, the failure of the image area extraction will also affect the normal operation of related applications, or bring functional use on the trouble. At this stage, there is no good way to detect the specific application of the image region extraction result. Cont...

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): G06K9/20G06K9/46
CPCG06V10/22G06V10/44
Inventor 徐爱辉
Owner NUBIA TECHNOLOGY CO LTD