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

Double transformation domain-based high magnification super resolution method

A super-resolution and high-resolution technology, applied in the field of computer vision, can solve problems such as unsatisfactory high-magnification reconstruction results

Active Publication Date: 2018-02-02
TIANJIN UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, although the double-transform domain method can achieve excellent results on the small-magnification super-resolution problem, the high-magnification reconstruction results are still unsatisfactory.

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
  • Double transformation domain-based high magnification super resolution method
  • Double transformation domain-based high magnification super resolution method
  • Double transformation domain-based high magnification super resolution method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0045] 1) Turn a high-magnification problem into a multi-level small-magnification cascading problem. For each level of small-magnification super-resolution problem, use formula (1) to model:

[0046]

[0047]

[0048] Among them, let F(α)=E D (x,y)+λ 1 E. TR (x,z)+λ 2 E. SR (x,z), x∈R n , y∈R m ,z∈R n Respectively, the vectorized restored depth map, the vectorized low-resolution depth map and the vectorized high-resolution color map, E D (x,y) is the data item, E TR (x,z) is a regular term in the transform domain, E SR (x,z) is the space regular term, λ 1 and lambda 2 is the weight parameter for balancing each item, and min means seeking the minimum value, Indicates the value of the variable x when the following formula is minimized; H indicates the synthesis operator of fuzzy and sampling, α i means block x i in the dictionary under the coefficient, means block x i The dictionary corresponding to the type of block, x i is the image block centered at...

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 relates to the field of computer vision and is used for improving depth map high magnification super resolution quantitative and qualitative results based on double transformation domains. According to a technical solution adopted in the invention, a double transformation domain-based multi-grade cascaded depth map super resolution method is employed and comprises the following steps: a high magnification problem is changed into multi-grade small magnification problems; for small magnification super resolution problems on each grade, modeling operation is performed via formula (1) described in the specification of the invention; final small magnification super resolution results can be obtained after the formula (1) is solved. The method disclosed in the invention is mainly applied to computer vision processing occasions.

Description

technical field [0001] The invention belongs to the field of computer vision. Regular terms based on frequency domain and space domain can obtain the latest level of results for small-magnification super-resolution, but cannot obtain satisfactory results for high-magnification super-resolution. The present invention makes innovative improvements to this problem. Background technique [0002] Depth information has been widely used in virtual reality, augmented reality, and 3D reconstruction. Despite the rapid development of depth cameras in recent years, the resolution of depth maps obtained by mainstream depth sensors (such as Kinect and TOF) is still far lower than that of color maps. Therefore, it is of practical significance to propose an effective depth map super-resolution method. At present, there are many excellent methods that have achieved good results on the small-magnification depth map super-resolution problem, but the high-magnification super-resolution proble...

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): G06T3/40
CPCG06T3/4053
Inventor 姜中玉杨敬钰岳焕景侯永宏侯春萍
Owner TIANJIN UNIV