Identification camera capable of automatically carrying out portrait cutout and method thereof

A camera and portrait technology, used in image enhancement, image analysis, image data processing, etc., can solve the problems of not meeting the requirements of portrait matting accuracy, mistakenly treating the background as a portrait, and consuming manpower and material resources.

Active Publication Date: 2015-12-23
GUANGZHOU XINGFU NETWORK TECH
View PDF4 Cites 27 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] ID photos are photos that meet the implementation standards of national legal documents such as ID cards, passports, Hong Kong travel permits, Macau travel permits, Taiwan travel permits, visas for going abroad, residence permits, and social security cards. It takes a lot of manpower and material resources to manually perform complex actions such as portrait range selection, cutout, and background replacement on the image on the processing software.
[0003] Although there are some APP software that can automatically change the background, but for pictures with a relatively messy background, the processing is relatively rough, especially at the edge of the avatar where the hair is cut, the processing is not smooth, and the background is mistakenly taken when the color of the background and the portrait are close. When it comes to portraits, it is acceptable only for entertainment, headshots, and resume photos, but for legal ID photos, it obviously cannot meet the requirements for the accuracy of portrait matting

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
  • Identification camera capable of automatically carrying out portrait cutout and method thereof
  • Identification camera capable of automatically carrying out portrait cutout and method thereof
  • Identification camera capable of automatically carrying out portrait cutout and method thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0069] The ID camera is an APP (Application, application program) installed on the smart terminal. It can have an Android version and an iOS version. The image data obtained by real-time shooting, the useful area of ​​the image is cropped according to the license standard, and the image information such as color, brightness, background, and biological characteristics of people has not been processed in any way.

[0070] However, the ID camera of the present embodiment is aimed at the original image of the photographed ID card, and performs matting processing on the original ID image to realize the separation of the foreground and the background (generally, the foreground refers to the portrait, and the background refers to the image other than the portrait. part), which includes a portrait contour search module, an initial foreground and initial background setting module, an initial foreground and expected portrait contour similarity detection module, an initial foreground and ...

Embodiment 2

[0098] Such as Figure 4 As shown, the present embodiment provides a method for automatically matting portraits, which is applied to ID cameras and includes the following steps:

[0099] S1. Use contour detection technology to quickly find the contour of the portrait in the original image of the license, including:

[0100] S11. Convert the original image of the certificate into a grayscale image;

[0101] S12. Perform upper body feature data area detection on the grayscale image, when the upper body feature data area is detected, save the area pixel information as an upper body area rectangle data structure, and set the upper body ROI area for the grayscale image, which is recorded as ROIB;

[0102] S13, ROIB is carried out face profile feature data area detection, when detecting the face profile feature data area, this area pixel information is saved as the face profile area rectangle data structure, and the face profile ROI area is set to the grayscale image, Recorded as ...

Embodiment 3

[0119] Such as Figure 5 As shown, the difference between the method for automatically matting portraits in this embodiment and embodiment 2 lies in step S8, as follows:

[0120] S8. Erase the pixel data on the original image mapped to the background area of ​​the final mask bitmap, and fill in a transparent color to obtain a foreground image.

[0121] To sum up, the ID camera and method of the present invention use the contour detection technology to quickly find the profile of the portrait in the original image of the license, and set the initial foreground area and the initial background area according to the profile of the found portrait, and use the mask bit of the initial foreground The mask bitmap of the image and the initial background, the mixed Gaussian model is established through the GrabCut function, and then the final mask bitmap of the foreground area and the background area is obtained through iterative adjustment. On the one hand, the speed of operation is imp...

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 identification camera capable of automatically carrying out portrait cutout and a method thereof. The identification camera comprises a portrait contour searching module, an initial foreground and initial background setting module, a foreground area and background area determination module and a foreground image separation module. The method comprises the following steps: carrying out quick searching on an original image by utilizing contour detection technology to obtain a portrait contour; setting the area in the found portrait contour to be the initial foreground, and setting the peripheral area of the found portrait contour to be the initial background; establishing a mixture Gaussian model through a GrabCut function by utilizing a mask bitmap of the initial foreground and a mask bitmap of the initial background, and then, obtaining the final mask bitmap of the initial foreground and the initial background through iteration adjustment; and establishing a transparent and empty image bitmap consistent with the original image in size, and copying pixels mapped in the foreground area of the final mask bitmap to the image bitmap to obtain a foreground image. The method can improve operation speed and cutout precision.

Description

technical field [0001] The invention relates to an ID camera and a method for automatically cutting out portraits, and belongs to the technical field of ID photo shooting and processing. Background technique [0002] ID photos are photos that meet the implementation standards of national legal documents such as ID cards, passports, Hong Kong travel permits, Macau travel permits, Taiwan travel permits, visas for going abroad, residence permits, and social security cards. It takes a lot of manpower and material resources to manually perform complex actions such as selecting the range of the portrait, cutting out the image, and changing the background on the image on the processing software. [0003] Although there are some APP software that can automatically change the background, but for pictures with a relatively messy background, the processing is relatively rough, especially at the edge of the avatar where the hair is cut, the processing is not smooth, and the background i...

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): G06T7/00
CPCG06T2207/20104
Inventor 不公告发明人
Owner GUANGZHOU XINGFU NETWORK TECH
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