A way to convert nv12+ format to p010 format

Abstract

The invention discloses a method for converting an NV12 + format into a P010 format. The method comprises the following steps:: 1) respectively moving 8-bit data of a YUV component and additional 2-bit data into two registers by using a VMOVDQA command; 2) moving the aligned packaging integer values by using a VPERMQ instruction, and averagely moving the data stored on the continuous address of the register to the high-low bit address of the AVX register; 3) interleaving and combining the data in the register with 0 by using a VPUNPCKLBW instruction, namely, data in a register and all 0 crosscombinations are placed in the register at intervals; and after the cross combinations, shifting data D in the NV12 format leftwards by 8 bits according to words, wherein the position of data C of additional 2-bit data is kept unchanged; and 4) sequentially finishing the addition of the register value of the D data for storing the NV12 and the register value of the C data for storing the additional 2 bits through the VPADDW instruction, and finishing the conversion from the NV12 + format to the P010be format. The method can quickly convert the NV12 + format into the P010 format.

Description

technical field [0001] The invention belongs to the field of image coding and relates to a method for converting an NV12+ format into a P010 format. Background technique [0002] Many hardware encoders (such as QSV, NVENC) natively support NV12 format image input, P010 storage format is a 10bit video image encoding format, its YUV adopts 4:2:0 sampling, and the storage order is to store all Y first, and then UV is stored alternately, and YUV is 10 bits. P010 is stored in 16 bits (2 bytes). According to the difference in size and size, it is divided into P010le and P010be. P010le places 10 bits in the lower 10 bits of 16 bits; P010be places 10 bits in the 10 bits higher than 16 bits. Its storage format is as figure 1 Shown: [0003] The 10-bit image stored by the P010 has 25% more information than the 8-bit image stored by the NV12, but has 100% more storage space. This can have an adverse effect on image acquisition. Therefore, a new image encoding method has emerged, c...

AI Technical Summary

Problems solved by technology

[0004] NV12+ is composed of additional data added by NV12. For 8-bit video encoding, NV12 data can be used directly; for 10-bit video encoding, if the NV12+ format can be quickly converted to P010 format, NV12+ can only store 10-bit image effective information without data redundancy. In addition, the data transmission efficiency of the acquisition card can be effectively improved, the storage space is reduced, and the data transmission efficiency of the acquisition card is improved. However, there is no similar format conversion method in the prior art

Images