Search

CN-122002021-A - Stereo video automatic separation and synchronous signal generation method and system based on embedded identification

CN122002021ACN 122002021 ACN122002021 ACN 122002021ACN-122002021-A

Abstract

The invention discloses a method and a system for automatically separating a stereoscopic video and generating a synchronous signal based on an embedded identification. Based on the original StereoMark scheme, syncMark synchronous marks are integrated, and a stereoscopic format description and synchronous parameters are simultaneously embedded in a video coding layer through double SEI messages, so that double backup is realized in a container layer. The playing end generates high-precision left and right frame synchronizing signals according to the synchronizing parameters, and outputs the signals through interfaces such as HDMI VRR, DP ADAPTIVE SYNC and the like, so that the synchronizing precision is smaller than 1 ms. And the self-adaptive network delay compensation is supported to be compatible with multiple display devices, so that the 3D viewing experience is remarkably improved. The method is suitable for scenes such as VR head display, vehicle-mounted 5D cabins, sports live broadcast and the like.

Inventors

  • YUAN GUANGMING
  • ZHAO XIAOMING

Assignees

  • 北京汉美奥科节能设备有限公司

Dates

Publication Date
20260508
Application Date
20260313

Claims (7)

  1. 1. The method for automatically separating the stereoscopic video and generating the synchronous signal is characterized by comprising the following steps of: Embedding a double SEI message in an SEI unit of a video coding frame, wherein the SEI message comprises a stereoo_mark_flag containing a stereoscopic format description and a sync_mark_flag containing a synchronization parameter; Storing the same dual identification data in a metadata box of the video container file; the playing end analyzes the SEI data preferentially, and retrieves the container metadata after failure; separating a single-frame video into a left view frame and a right view frame according to the layout parameters analyzed by the sterio_mark_flag; And generating a left frame synchronization signal and a right frame synchronization signal according to the sync mark flag, and outputting the signals to the display equipment through a special interface or an embedding mode.
  2. 2. The method of claim 1, wherein the sync_mark_flag comprises: frame_number: global frame number; timestamp_us: microsecond level timestamp; left_view_delay_us-right_view_delay_us: a left-right view delay compensation value; sync_ accuracy _us, synchronization precision index; display_ready_signal, display ready signal flag; adaptive_sync_support: list of supported synchronization interfaces.
  3. 3. The method of claim 1, further comprising the step of adaptive synchronization compensation: calculating a left-right view arrival time difference delta t through network RTT measurement; Dynamically adjusting the depth of the left and right view buffer queues to eliminate deltat; The PLL phase locked loop algorithm is used to maintain the synchronization error less than 1ms.
  4. 4. The method of claim 1, wherein the synchronization signal output means comprises: HDMI 2.1 VRR signal, transmitting synchronous pulse through VRR pin; DisplayPort ADAPTIVE SYNC, transmitting the synchronization parameters through the AUX channel; inserting a synchronous code pattern in a frame blanking period; And an independent audio channel, which is used for transmitting synchronous information through audio carrier modulation.
  5. 5. A stereoscopic video playback system, comprising: A dual identity embedding module (performing dual SEI embedding of right 1); A synchronization signal processing module (performing the synchronization signal generation of the right 1); An adaptive synchronous compensation module (performing delay compensation of weight 3); multiple interfaces synchronous output module (various output modes supporting right 4).
  6. 6. The system of claim 5, further comprising an on-board 5D linkage controller: The seat vibration, atmosphere lamp and fragrance system are linked after the synchronous signals are received; The somatosensory effect is ensured to be synchronized with the 3D picture (error <5 ms).
  7. 7. The system of claim 5, wherein the multi-interface synchronous output module comprises: the VR split screen synchronizer supports OpenXR synchronous APIs; the network synchronization adapter supports WebRTC synchronization expansion; and the wireless screen-throwing synchronizer supports Miracast 2.0 synchronization protocol.

Description

Stereo video automatic separation and synchronous signal generation method and system based on embedded identification Technical Field The invention relates to the technical field of video encoding and decoding and multimedia processing, in particular to a method and a system for embedding separable marks in video signals and generating left and right frame synchronous signals, which are suitable for 3D video playing, virtual Reality (VR), augmented Reality (AR), multi-view interaction and active shutter type 3D display scenes. Background Pain point in the prior art: the hard coding format depends on that the current stereo video needs to be in a preset format, and the playing end needs to manually configure decoding parameters. The lack of synchronization signal, which is that the existing stereoscopic video format (SBS/TB) only contains image data, lacks frame-level synchronization signal, causes the left and right eye pictures to shift on the time axis (typical error is 5-15 ms), and causes 3D dizziness (IEEE VR 2023 research report). The mark is vulnerable to visible watermark damage composition and steganography is easily damaged by clipping/compression. Transmission delays are inconsistent-left and right views experience different path delays in network transmission, lacking a compensation mechanism (RFC 7273). Related patent limitations: US20180034821A1 (sony) only solves format identification, and does not involve synchronization problems. CN113114658a (hua) relies on special hardware and has poor versatility. EP2896278A1 (samsung) uses external synchronization lines, which are not suitable for wireless/network transmission scenarios. In conclusion, a stereoscopic video processing scheme integrating synchronous signal generation, adaptive delay compensation and cross-platform synchronization is needed. Disclosure of Invention 1. Technical problem to be solved Eliminating manual format configuration during stereoscopic video playing; Solving the problem of left and right frame time synchronization offset (target accuracy <1 ms); The asynchronous left and right views caused by network transmission is avoided; Synchronous interfaces (HDMI 2.1 VRR, displayPort ADAPTIVE SYNC, wireless screen casting) supporting multiple 3D display devices. 2. Summary of the technical solution Based on the original StereoMark scheme, a SyncMark synchronous marking system is added, and the core comprises: Triple identification embedding system: coding layer (SEI) inserting a double SEI message in H.265/AV 1: stereoformat description (layout type, resolution ratio) Sync_mark_flag synchronization signal data (time stamp, frame number, delay offset value) The container layer (Metadata Box) stores the same data in the uuid Box of MP4 The synchronous signal layer (special channel) transmits accurate synchronous pulse via HDMI VRR/DP ADAPTIVE SYNC interface SyncMark synchronization signal design: { "sync_mark_version": "1.0", "frame_number": 12345, "timestamp_us": 1703856789012345, "left_view_delay_us": 0, "right_view_delay_us": 850, "sync_accuracy_us": 150, "display_ready_signal": true, "adaptive_sync_support": ["HDMI2.1_VRR", "DP_ASYNC"] } A playing end synchronization engine: graph TD A [ video input ] - - - > B { SyncMark detect } B- -detected- - > C [ resolution of stereoscopic format+synchronization parameter ] B- -no detection- - > D [ play in 2D ] C- > E [ decoding video frame ] E- > F [ view separation Engine ] F- > G [ synchronous signal processing Module ] G- > H { delay compensation calculation } H- > |network delay variance|I [ left view buffer queue ] H- > |network delay variance |J [ Right buffer queue ] I- > K [ synchronous output controller ] J -->K K- > L [3D rendering+synchronization Signal output ] Adaptive synchronization compensation algorithm: Network delay measurement, counting the left and right view arrival time difference delta t through RTCP feedback packets. Buffer dynamic adjustment the left and right view buffer depth (range: 0-50 ms) is dynamically adjusted according to Δt. Synchronization accuracy calibration-maintaining a synchronization error of <1ms using a PLL phase-locked loop algorithm. 3. Advantageous effects High precision synchronization-left and right frame synchronization errors are reduced from 10-15ms of the traditional scheme to <1ms (measured data: 1000 frame samples). The user experience is improved, and the occurrence rate of the 3D dizziness symptom is reduced by 78 percent (based on 500 user tests). The compatibility is enhanced, and a plurality of synchronous interfaces such as HDMI 2.1 VRR, displayPort ADAPTIVE SYNC, USB-C Alt Mode and the like are supported. Network adaptation-synchronization accuracy <2ms can be maintained in a network environment with rtt=50 ms. Commercial value, has been applied to the video 5D cabin system of miaow, realize on-vehicle 3D and watch shadow zero dizziness. Drawings FIG. 1 is a diagram of a StereoMark system architecture for integration Sync