Video super-resolution reconstruction method and system based on temporal feature fusion
A technology of super-resolution reconstruction and temporal features, which is applied in the field of video super-resolution reconstruction based on temporal feature fusion, can solve the problems of missing image details, ignoring global features, and poor super-resolution reconstruction effects, etc. The effect of the reconstruction effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0035] Aiming at the problem that the previous video super-resolution reconstruction technology based on convolutional neural network fails to effectively utilize the local and global information in the video, which leads to poor video reconstruction quality, this embodiment proposes a method based on temporal feature fusion. The video super-resolution reconstruction method uses complementary features from the global time domain to improve the super-resolution reconstruction effect of each frame in the video sequence while filtering out the effective features in the local time domain.
[0036] refer to figure 1 , this embodiment provides a video super-resolution reconstruction method based on temporal feature fusion, which specifically includes the following steps:
[0037] S101: Acquire an image sequence of a video, extract features of the image sequence, and obtain an initial feature sequence.
[0038] In a specific implementation, for a set resolution video image sequence ...
Embodiment 2
[0047] This embodiment provides a video super-resolution reconstruction system based on time-domain feature fusion, which specifically includes the following modules:
[0048] an initial feature extraction module, which is used to obtain the image sequence of the video, extract the features of the image sequence, and obtain the initial feature sequence;
[0049] The local feature fusion module is used to fuse the features in the initial feature sequence with local time domain features to obtain a local feature sequence; wherein, the non-boundary features in the initial feature sequence are fused with their two nearest neighbors; for the initial feature sequence For the boundary features in the feature sequence, two of the boundary features and the one closest to them are fused;
[0050] The global feature fusion module is used to input the local feature sequence into the variable convolutional short-term memory network of bidirectional sampling, and perform global feature supp...
Embodiment 3
[0054] This embodiment provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, implements the method in the video super-resolution reconstruction method based on temporal feature fusion described in the first embodiment above. step.
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More