Light supplementing method and system for human face in image, medium and electronic equipment
A technology for filling light in images, applied in the field of image processing, can solve problems affecting user experience and poor light filling effect, achieve excellent light filling effect, improve user experience, and avoid the risk of background overexposure
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0078] This embodiment provides a method for supplementing light on a human face in an image. The image may be a static picture or a dynamic video including a background and a portrait. This embodiment does not limit the specific form of the image.
[0079] The supplementary light refers to adding light effects such as highlights, shadows, and spot lighting to the human face, so as to realize the human face photography effect of simulating artificial lighting in a studio, and improve the beauty of the human face.
[0080] Such as figure 1 As shown, the supplementary light method may include the following steps:
[0081] Step S10: identifying the human face in the image;
[0082] Step S11: Obtain the 3D dense key points of the face;
[0083] In this embodiment, automatic face recognition and three-dimensional face dense key point alignment technology can be used to accurately locate the position of the face in real time, and fit the pose, expression and shape of the face.
...
Embodiment 2
[0106] This embodiment provides a method for supplementing light for a human face in an image, and the method for supplementing light is a further improvement on the basis of Embodiment 1.
[0107] Further, see figure 2 , before the step S11 may also include:
[0108] Step S20: Acquiring the number of faces in the image;
[0109] Step S21: judging whether the number of faces is 1;
[0110] If the number is 1, then perform step S22: calculate the distance between the real face and the imaging device during imaging;
[0111] Step S23: judging whether the distance does not exceed a preset distance threshold;
[0112] If the preset distance threshold is not exceeded, step S11 is executed.
[0113] In this embodiment, considering that the multi-face scene is relatively complex and cross-occlusion of human faces may occur, if the number of human faces in the detected image is more than 2, no supplementary lighting is performed on the human faces in the image. Similarly, if the...
Embodiment 3
[0128] This embodiment provides a light supplement method for a human face in an image, which is a further improvement on the basis of Embodiment 1.
[0129] Further, as image 3 As shown, before the step S14, it may also include;
[0130] Step S30: performing spatial triangulation on the three-dimensional dense key points to obtain an array of triangles, each of which includes three three-dimensional dense key points;
[0131] In the process of building triangles, multiple triangles can share vertices (ie, key points), but the plane of each triangle does not intersect with the plane of any other triangle.
[0132] Preferably, a C program (a programming language) can be used to implement the steps in this algorithm.
[0133] For example: dense key points can be expressed as Verts[cout][3] in a programmatic statement, and spatial triangulation is performed on three-dimensional dense key points to obtain a triangle array Triangles[num][3].
[0134] Among them, cout is the num...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com