Video time-space super-resolution reconstruction method based on robust optical flow and Zernike invariant moment

Abstract

The invention discloses a video time-space super-resolution reconstruction method based on a robust optical flow and Zernike invariant moment. The method comprises the following steps: performing motion analysis on a video sequence in a time-space domain, constructing a robust optical flow motion estimation model of the video sequence, and obtaining a motion vector; according to the motion vector, performing bidirectional time-space motion compensation on the video sequence to obtain a compensated video sequence; and by use of a cross-scale fusion strategy of a Zernike invariant moment based rapid non-local fuzzy registering mechanism, performing time-space super-resolution reconstruction on the compensated video sequence to obtain a video sequence with high time-space resolution. According to the invention, the method is not dependent on accurate sub-pixel motion estimation, can be applied to various complex motion modes, such as angle rotation, local motion and the like, can provide clear and smooth video information for accurate identification and tracking of a motion object, and has a quite high actual application value.

Description

technical field [0001] The invention relates to the technical field of image / video enhancement, in particular to a video space-time super-resolution reconstruction method based on robust optical flow and Zernike invariant moments. Background technique [0002] Resolution is an important indicator of the quality of spatial moving images. The higher the resolution, the more detailed information about spatial moving objects obtained from the moving image sequence, which is more conducive to the precise identification and tracking of spatial moving objects. Due to factors such as motion or optical blur, undersampling and noise interference, the visual effect of moving image sequences is poor. [0003] Traditional super-resolution reconstruction methods rely on precise sub-pixel motion estimation and thus are limited to some simple motion patterns such as global translation. There are some complex motion patterns in the video sequence. When there are rotations at different angle...

AI Technical Summary

Problems solved by technology

There are some complex motion patterns in the video sequence. When there are rotations at different angles in the video sequence, the spatiotemporal similarity between frames becomes very weak, and it is difficult to effectively use the information between frames, which in turn affects super-resolution. quality of reconstruction

The traditional frame interpolation technology based on motion vector is inevitably affected by motion estimation error, which will cause visual block effect or hole effect in the interpolated frame

[0004] The iterative back-projection method, the maximum a posteriori estimation MAP, and the convex set projection method POCS often rely on accurate sub-pixel motion estimation, and can only achieve better results in some simple motion modes such as global translation and a single moving target scene. effect, it cannot effectively handle some complex motion scenes, and it is difficult to achieve accurate motion estimation, which seriously affects the quality of super-resolution reconstruction

Images