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

Image warping method based on 3D saliency

An image distortion and significant technology, applied in the field of image processing, can solve problems such as image distortion, loss of important information, object deformation, etc., to achieve strong noise resistance and ensure robustness

Active Publication Date: 2017-05-17
HEFEI UNIV OF TECH
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention aims to solve the problems that most current image distortion methods will deform objects in the image, lose some important information and cannot effectively distort images with complex scenes during the process of distorting the target image. 3D saliency image warping method, which can combine depth information with 2D saliency and enhance the robustness of distortion

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 warping method based on 3D saliency
  • Image warping method based on 3D saliency
  • Image warping method based on 3D saliency

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0046] In this embodiment, an image distortion method based on three-dimensional saliency is to use depth data to obtain a depth map of the target image; then combine the depth map and the two-dimensional model to construct a three-dimensional saliency model; Adaptively update the weight between the depth data and the 2D model; use the 3D saliency model again to calculate the gradient of the image energy function; then extract the 2D edge and depth profile features of the image, and use the feature points to generate a triangular mesh; finally construct the objective function to find Take the extremum and add constraints to get the distortion matrix; specifically, follow the steps below:

[0047]Step 1: Use formula (1) to calculate the energy function E of each pixel in the target image I with an image size of m×n: the energy function is obtained by using the L1 paradigm to calculate the gradient of each pixel of the target image; the gradient information It usually reflects t...

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 distortion method based on three-dimensional saliency, which is characterized by following the following steps: 1. using depth data to obtain a depth map of a target image; 2. combining the depth map and a two-dimensional model to build a three-dimensional saliency model; 3. Adaptively update the weight between the depth data and the two-dimensional model according to the distribution of image grayscale; 4. Use the three-dimensional saliency model to calculate the image energy function gradient; 5. Extract the two-dimensional edge and depth contour features of the image, and use feature points to generate triangles Grid; 6. Construct the objective equation and add constraints; 7. Find the extreme value of the objective equation and construct the transformation relationship. The invention can combine depth information with two-dimensional saliency and enhance the robustness of distortion.

Description

technical field [0001] The invention belongs to the field of image processing, and mainly relates to an image warping method based on three-dimensional saliency. Background technique [0002] With the popularity of mobile devices such as smart phones and tablet computers, users are accustomed to uploading the photos they have taken to social networking sites to share with friends. Considering the different models of mobile devices used by users, how to make the photos shared by users can be well displayed on different terminals is one of the hot topics in the field of computer vision research. [0003] In response to the above problems, researchers have proposed some methods, but due to distortion and insufficient detection accuracy, the deformation of objects in the image and the loss of some important information have always plagued researchers. Recently, the article "Saliency detection for content-aware image resizing" was published on the IEEE International Conference o...

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 Patents(China)
IPC IPC(8): G06T3/00
CPCG06T7/593G06T2207/10028G06T3/067
Inventor 汪萌高欣健陈雁翔潘宜飞
Owner HEFEI UNIV OF TECH