Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

An Interpolation Method for Non-rigid Image Sequence

An image sequence and inter-frame interpolation technology, applied in image analysis, image enhancement, image communication, etc., can solve the problems of inter-frame interpolation error, difficult to obtain accurate dense vector field, difficult to obtain accurate optical flow field, etc., to achieve The effect of high accuracy

Active Publication Date: 2021-11-09
TIANJIN UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Existing interpolation methods are almost all interpolated for optical flow fields with only a small amount of missing data, and are not suitable for vector fields with only a small amount of sparse data
Moreover, there are large local displacements and serious nonlinear deformations in non-rigid images, and it is difficult to obtain accurate optical flow fields.
[0005] Therefore, the above method is inevitably limited by the inaccuracy of the optical flow field, and it is difficult to obtain an accurate dense vector field after interpolation, resulting in large errors in inter-frame interpolation

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
  • An Interpolation Method for Non-rigid Image Sequence
  • An Interpolation Method for Non-rigid Image Sequence
  • An Interpolation Method for Non-rigid Image Sequence

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0040] The embodiment of the present invention provides an interpolation method based on sparse vector field modeling, which mainly consists of five parts: selection and matching of sparse feature points; front-to-back consistency detection to remove false matching; modeling interpolation of sparse vector field; Estimate dense vector fields between frames using spline interpolation; correct floating images based on dense vector fields.

[0041] The embodiment of the present invention realizes the modeling of the dense vector field based on a small number of sparse non-uniform vector fields, and is not limited to the inaccuracy of the sparse optical flow field. Not only can full-pixel registration of non-rigid images with large displacements be realized, but also inter-frame interpolation can be realized based on a small number of non-rigid image sequences. Such as figure 1 As shown, the specific steps and principles are as follows:

[0042] 101: Use robust local optical flow...

Embodiment 2

[0083] The scheme in embodiment 1 is described in further detail below in conjunction with specific examples, see the following description for details:

[0084] Figure 5 (a) with Figure 5 There are non-rigid deformations with large displacements between the flower images in (b). After the initial matching and the consistency check, the accurate sparse vector field can be established by using the correct matching points, such as Figure 5 (c), where the arrows represent the feature points from Figure 5 (a) Position orientation Figure 5 (b) A vector of positions. The dense vector field obtained by interpolation using this method is as follows Figure 5 As shown in (d), for the convenience of observation, the dense displacement vector field is down-sampled by 15 times. Depend on Figure 5 (d) It can be seen that this method can effectively obtain a dense vector field from a sparse vector field.

[0085] Image 6 (a) and (b) are the template image and the floating ima...

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 inter-frame interpolation method for a non-rigid image sequence. The method includes: using robust local optical flow to initially match sparse feature points composed of feature points on moving edges and uniform grid points ;Remove the mismatching feature points through the front-back consistency detection criterion, obtain sparse matching points, and then obtain the sparse vector field; perform modeling interpolation on the sparse vector field to obtain a dense motion vector field; use the spline interpolation method to calculate the The dense vector field in between is estimated, and the floating image is corrected according to the dense vector field, so that more continuous frames are interpolated from a few frames. The present invention realizes the modeling of the dense vector field based on a small amount of sparse non-uniform vector field, and is not limited to the inaccuracy of the sparse optical flow field. Non-rigid image interpolation generates multiple consecutive image sequences.

Description

technical field [0001] The invention relates to the field of computer vision inter-frame interpolation, in particular to an inter-frame interpolation method for non-rigid image sequences, which can be used for registration of non-rigid deformed images and to obtain multiple continuous image sequences based on a small number of images. Background technique [0002] Interpolation of non-rigid image sequences is a challenging problem in computer vision. Due to the extensive local large displacement deformation and severe nonlinear deformation between non-rigid image sequences, it is difficult to accurately obtain the dense displacement vector field between any two images, and it is even more difficult to obtain the dense displacement vector field between consecutive frames . The inter-frame interpolation of the sequence is mainly realized by estimating the dense vector field between frames, the purpose is to generate an unknown intermediate frame between any two consecutive or...

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): H04N19/587G06T7/246G06T7/269
CPCG06T2207/10016G06T7/248G06T7/269H04N19/587
Inventor 何凯王阳刘坤冯旭
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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