Video recording medium and video coding method
A recording medium and coding technology, which is applied in the direction of digital video signal modification, TV, color TV, etc., can solve the problems of large residual error and reduced coding efficiency, achieve good concealment, avoid coding efficiency, and the code stream structure is not affected. Effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0048] This embodiment describes the implementation process of the present invention in a video codec using a block matching algorithm.
[0049] Firstly, the block matching algorithm in motion estimation is introduced. In video coding, the current frame is divided into several non-overlapping regions, such as macroblocks. All pixels in the same macroblock are assumed to have the same displacement. Each motion estimation operation involves at least two images, the current image and the reference image. In order to describe the movement of the content of the macroblock relative to the previous image (coding order), search for blocks of the same size in the reference frame in a certain area around the same position of the macroblock of the current image, and find a block with the same size as the macroblock of the current image based on the specified matching criteria. The block closest to the block is used as the matching macroblock. The same operation is performed on all mac...
Embodiment 2
[0062] A premise in Embodiment 1 is that the location of the macroblock where the watermark information is embedded is known to both the watermark embedding unit of the encoder and the watermark extraction unit of the decoder. The following embodiment 2 will illustrate how to realize adding and extracting watermark without this premise.
[0063] On the encoder side, with Figure 7 The steps shown for adding a watermark are unchanged. However, watermarks can be embedded in any macroblock. Moreover, since the embedding position of the watermark is unknown, it is necessary to be able to include the information in the watermark, or to distinguish the macroblock embedded with the watermark from the macroblock not embedded. One method is to reduce the effective storage bit by one bit, that is, the watermark information must be selected from a location point other than the minimum error location point. The macroblock without watermark is identified by the motion vector as the mini...
Embodiment 3
[0067] The above embedding process assumes that the difference between the reconstructed image and the input image to be encoded by the encoder is negligible, that is, because the quantization error will not cause the encoder and decoder to have an impact on watermark embedding and extraction. This embodiment illustrates the watermark embedding and extraction method when the quantization error is large enough.
[0068] Quantization error is produced by the quantizer. The quantizer is used as Figure 13 The quantization steps shown represent the input data with a larger quantization step. Although the accuracy is reduced compared to the original data, the number of bits representing these data is also greatly reduced. For example, the pre-quantized numbers between 5 and 15 will be quantized to 4, so that the decimal part does not need to be encoded again. During inverse quantization, the number 4 is converted into a number between 5 and 15, for example, 10, and 10 is the inpu...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More