Frequency domain registration and convex set projection-based multi-frame image super-resolution reconstruction method
A technology for super-resolution reconstruction and multi-frame images, which is applied in image enhancement, image data processing, instruments, etc. It can solve the problems of low-resolution images, functioning, extracting and analyzing single-frame images, etc.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment
[0061] Example: see figure 1 , figure 2 . This multi-frame image super-resolution reconstruction method based on frequency domain registration and convex set projection is characterized in that its specific implementation steps are:
[0062] A. Read in multiple frames of images and select one as a reference frame for registration
[0063] B. Image registration: convert the image to the frequency domain, and in the case of a small offset, only estimate the horizontal and vertical displacement components, ignoring the rotation angle;
[0064] C. Image reconstruction: according to the requested displacement parameters, register the low-resolution image to the high-resolution grid, and perform super-resolution reconstruction using the convex set projection algorithm;
[0065] D. Output image
Embodiment 2
[0067] This embodiment is basically the same as Embodiment 1, and the special features are as follows:
[0068] This embodiment first extracts 10 consecutive frames of a video sequence, respectively f 1 (x), f 2 (x)... f 10 (x), the size is 90×90. like Figure 5 Shown is one of ten images selected from the video sequence.
[0069] (1) Image Registration
[0070] (1) For the input image f 1 (x), f 2 (x)... f 9 (x) perform Fourier transform, and select f 1 (x) is used as a reference image, and the others are images for which considerable displacements are to be estimated. The Fourier transform result is F 1 (u)F 2 (u)...F 10 (u).
[0071] (2) According to εU max s -U max Select the spectral range of the image, which is usually in the center third of the image. like Figure 5 shown in blue.
[0072] (3) Calculate the relative displacement between images according to formula 4, using the least squares method.
[0073] The translation parameters of the reference ...
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