Backward Motion Estimation Research Articles

Advanced Patch-Based Affine Motion Estimation for Dynamic Point Cloud Geometry Compression.

The substantial data volume within dynamic point clouds representing three-dimensional moving entities necessitates advancements in compression techniques. Motion estimation (ME) is crucial for reducing point cloud temporal redundancy. Standard block-based ME schemes, which typically utilize the previously decoded point clouds as inter-reference frames, often yield inaccurate and translation-only estimates for dynamic point clouds. To overcome this limitation, we propose an advanced patch-based affine ME scheme for dynamic point cloud geometry compression. Our approach employs a forward-backward jointing ME strategy, generating affine motion-compensated frames for improved inter-geometry references. Before the forward ME process, point cloud motion analysis is conducted on previous frames to perceive motion characteristics. Then, a point cloud is segmented into deformable patches based on geometry correlation and motion coherence. During the forward ME process, affine motion models are introduced to depict the deformable patch motions from the reference to the current frame. Later, affine motion-compensated frames are exploited in the backward ME process to obtain refined motions for better coding performance. Experimental results demonstrate the superiority of our proposed scheme, achieving an average 6.28% geometry bitrate gain over the inter codec anchor. Additional results also validate the effectiveness of key modules within the proposed ME scheme.

Hierarchical Motion-Compensated Frame Rate Up-Conversion Using Gaussian Pyramid and Guidance Motion Vector

This paper proposes fast and accurate frame rate up-conversion algorithm for both software and hardware. The proposed algorithm first construct Gaussian pyramid structure using 3×3 Gaussian mask. After constructing Gaussian pyramid, forward and backward unilateral motion estimation is implemented at the top level of the pyramid. Then, motion vector field that is obtained at the top level is projected onto the middle level and the bottom level of the pyramid with unilateral motion vector refinement. Next, on the bottom level, we implement motion vector smoothing and proposed motion vector shifting with bilateral motion vector refinement. By shifting unilateral motion vector to bilateral motion vector, we could avoid generating holes and overlapped regions. Finally, we select final MV between forward and backward MVs according to their reliability. After all the processes, we interpolate new frames by using overlapped block motion compensation. Experimental results show that proposed algorithm is up to 80 times faster than conventional algorithm with 0.1dB peak signal-to-noise gain in average.

순방향과 역방향 움직임 추정을 이용한 효율적인 움직임 보상 외삽 기법

Motion compensated extrapolation (MCE) techniques show inferior performance compared to motion compensated interpolation techniques, since only past frames are used in MCE. MCE techniques are used for the reconstruction of corrupted frames, the up-conversion of frame rates and the generation of side information in the distributed video coding system. In this paper, the performance of various MCE techniques are evaluated and an efficient MCE technique using the forward and backward motion estimation is proposed. In the proposed technique, the present frame is extrapolated by averaging two frames which are generated by forward and backward motion estimation respectively. It is shown that the proposed method produces better PSNR results and less blocking phenomena than conventional methods.

A Five-Field Motion Compensated Deinterlacing Method Based on Vertical Motion

A new motion compensated deinterlacing method using forward and backward motion estimation is proposed in this paper. Bi-directional motion estimation is performed using two previous and two subsequent fields. The motion estimator uses pre-filtering prior to motion estimation for the current and the subsequent two fields. The motion estimator finds a single optimal matching block in the same or opposite parity reference fields. Motion compensation is done according to the amount of vertical motion within the reference fields to achieve the highest vertical resolution improvement. A novel technique to prevent the appearance of visual artifacts in the presence of fast-moving objects is proposed. Experimental results show that the proposed method performs better than the conventional deinterlacing methods, based on objective and subjective criteria.

Very low bit video coding algorithm using uncovered region memory

In order to solve problems of uncovered regions in region-based motion estimation, we address a novel method which generates the uncovered region memory using backward motion estimation, and present results of the application to very low bit rate video coding. In the proposed algorithm, changed regions are detected by using frame difference (FD) and segment information. For the changed region of a previously reconstructed frame, backward motion compensated prediction is performed. Then, with motion information, the uncovered region memory is generated and updated without transmission of shape information. Whenever the uncovered region appears in the scene, the algorithm adaptively selects between spatial prediction from the previously reconstructed image and temporal prediction from a special uncovered memory. The regions with large prediction error are transformed and encoded. Simulation results show an improvement in the subjective and objective image quality due to the remarkable reduction of transmission bits for prediction error.