NVIDIA DRIVE OS Linux SDK API Reference

5.2.6 Release
For Test and Development only
NvMediaEncodeMVBufferHeader Struct Reference

Detailed Description

Header format that defines motion vector output.

This header will be present in the encoded bitstream output at an offset defined by NvMediaEncodeOutputExtradata::MVBufferDumpStartOffset if NvMediaEncodeOutputExtradata::bMVbufferdump is set. The motion vector output will immediately follow this header if buffersize is non-zero.

Note
Motion Vector output format in memory (v2.0):
<Memory Address - low>
|-----------------------------| <- MVBufferDumpStartOffset
|-----------------------------|
|-----------------------------|
|-----------------------------|
.
.
<Memory Address - high>
To enable v2.0 format of the MV Buffer dump, NVMEDIA_ENCODE_CONFIG_H264_ENABLE_MV_BUFFER_DUMP_V2 should be bit-ORd in NvMediaEncodeConfigH264::features

Definition at line 160 of file nvmedia_iep_output_extradata.h.

Data Fields

uint32_t MagicNum
 Used to verify the integrity of the header. More...
 
uint32_t buffersize
 Size of motion vector output (excluding header size), i.e., the size of MV data in the bitstream post the header. More...
 
uint16_t blocksize
 Macro Block size. More...
 
uint16_t width_in_blocksize
 Input frame width in terms of blocksize. More...
 
uint16_t height_in_blocksize
 Input frame height in terms of blocksize. More...
 
uint16_t reserved
 Reserved. More...
 

Field Documentation

◆ blocksize

uint16_t NvMediaEncodeMVBufferHeader::blocksize

Macro Block size.

Definition at line 169 of file nvmedia_iep_output_extradata.h.

◆ buffersize

uint32_t NvMediaEncodeMVBufferHeader::buffersize

Size of motion vector output (excluding header size), i.e., the size of MV data in the bitstream post the header.

Definition at line 167 of file nvmedia_iep_output_extradata.h.

◆ height_in_blocksize

uint16_t NvMediaEncodeMVBufferHeader::height_in_blocksize

Input frame height in terms of blocksize.

Definition at line 173 of file nvmedia_iep_output_extradata.h.

◆ MagicNum

uint32_t NvMediaEncodeMVBufferHeader::MagicNum

Used to verify the integrity of the header.

IEP will set this to MV_BUFFER_HEADER to indicate that the header is valid.

Definition at line 164 of file nvmedia_iep_output_extradata.h.

◆ reserved

uint16_t NvMediaEncodeMVBufferHeader::reserved

Reserved.

Definition at line 175 of file nvmedia_iep_output_extradata.h.

◆ width_in_blocksize

uint16_t NvMediaEncodeMVBufferHeader::width_in_blocksize

Input frame width in terms of blocksize.

Definition at line 171 of file nvmedia_iep_output_extradata.h.


The documentation for this struct was generated from the following file:
NvMediaEncodeMVBufferHeader::width_in_blocksize
uint16_t width_in_blocksize
Input frame width in terms of blocksize.
Definition: nvmedia_iep_output_extradata.h:171
NvMediaEncodeMVBufferHeader::height_in_blocksize
uint16_t height_in_blocksize
Input frame height in terms of blocksize.
Definition: nvmedia_iep_output_extradata.h:173
NvMediaEncodeMVBufferHeader
Header format that defines motion vector output.
Definition: nvmedia_iep_output_extradata.h:160
NvMediaEncodeMVData
Motion Vector format - motion vectors for each of the macro blocks are dumped in this format contiguo...
Definition: nvmedia_iep_output_extradata.h:182