B-frame expansion direct mode oriented to panoramic video coding

Abstract

The invention discloses a B-frame expansion direct mode oriented to panoramic video coding. The mode comprises three expansion direct modes, i.e. (1) mode A: using the backward motion vector of a corresponding block in a forward reference frame to carry out time-domain scale to obtain the forward and backward motion vectors of the current block; (2) mode B: using the inverse mapping of the backward motion vector of a corresponding block in a backward reference frame to generate a forward motion vector, utilizing the mapped forward motion vector to scale to obtain the forward and backward motion vectors of the current block; and (3) mode C: using the forward motion vector of the corresponding block in the backward reference frame to carry out time-domain scale to obtain the forward and backward motion vectors of the current block. The invention has the advantages of effectively compensating the defects of the prior art, reducing the prediction residual and increasing the probability of selecting the direct mode, thus greatly saving the bit numbers for coding motion vectors and improving the coding performance of panoramic video sequences.

Description

technical field [0001] The invention relates to an image video coding compression technology, in particular to a B-frame direct mode method for panoramic video coding. Background technique [0002] Panoramic video is a kind of omnidirectional video based on image rendering, also known as immersive video, the purpose is to provide users with a virtual reality experience, so that users can observe the surrounding scenes stitched together with photo images. A panoramic image is a six-sided or eight-sided fisheye camera that rotates or zooms around a fixed axis, takes pictures of different orientations of the scene around the camera at the same time, and uses some "stitching" techniques to seamlessly combine these pictures. Seam splicing, and then mapped into a cylindrical or spherical map according to the cylindrical or spherical mapping algorithm. Therefore, the resolution of the panoramic image is generally higher (because it is stitched from multiple conventional images), a...

AI Technical Summary

Problems solved by technology

[0008] However, in previous standards (H.263 and MPEG-2), only I-frames and P-frames can be used as reference frames for prediction frames, and I-frames use an intra-frame prediction mode similar to JPEG2000, which does not generate motion Vector; while the P frame uses the forward prediction mode, only one forward motion vector is generated

Therefore, in the existing B frame direct mode, including in the latest standard H.264 / AVC, if the backward reference frame is a P frame, it can only be done using the forward motion vector of the corresponding block on the backward reference frame. Scaling in the time domain achieves the purpose of obtaining the forward and backward motion vector of the current block. In some scenes, the backward progressive video coding will obtain a relatively ideal coding effect, but in some scenes, the video with strong temporal correlation in the forward direction In encoding, such as panoramic video encoding, the direct mode is not available due to the lack of backward motion information

Furthermore, if the backward reference frame is encoded as an I frame, the front and rear motion vectors in the direct mode used by the current block to be encoded are all set to zero. Obviously, this will also greatly reduce the motion prediction accuracy of the direct mode

Images