Method for detecting all-zero DCT coefficients ahead of discrete cosine transformation and quantisation

Zhou, Xuan; Yu, Zhen; Yu, Songyu

doi:10.1049/el:19981308

Cited by 71 publications

(38 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Examples include row-based SAD [2] and the maximum value of the sum of positive residuals and the absolute value of negative residuals [10]. Methods for deriving the thresholds can be roughly classified into three categories: upper boundbased [7,9,10,16], distribution-based [8,17], and hybrid methods that employ multiple thresholds derived using methods in the first two categories [5,18].…”

Section: Related Workmentioning

confidence: 99%

“…These methods estimate the upper bound of a DCT coefficient F(u, v) as a product of a scalar α and the SAD [7,9,10]:…”

Section: Related Workmentioning

confidence: 99%

“…Among these methods, Zhou et al [7] provided a larger value α (i.e., 1 4 for an 8 × 8 DCT), which provided a more stringent condition for detecting AZBs. Sousa [9] refined the model proposed by Zhou et al and derived a more precise α (i.e., α = 1 4 cos 2 π 16 ).…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

A two-stage algorithm for the early detection of zero-quantized discrete cosine transform coefficients in High Efficiency Video Coding

Chen

Wang

Tian

2017

J Image Video Proc.

View full text Add to dashboard Cite

For High Efficiency Video Coding (HEVC) using block-based prediction, the discrete cosine transform (DCT), and quantization, a large number of DCT coefficients in a transform block (TB) are commonly found to be quantized to zero. A two-dimensional transform in HEVC is usually implemented by first applying a butterfly-based one-dimensional (1D) DCT to each row of the residual block, followed by a 1D DCT to each column. Accordingly, we propose a two-stage method for the early detection of zero-quantized DCT coefficients in this paper. In the first stage, a distribution-based method, which uses the sum of absolute differences as the threshold criterion and compares it to threshold values obtained for all columns, is employed to detect zero-quantized columns in the pixel domain prior to conducting the actual DCT. After conducting the 1D row DCT, the second detection stage, which uses the intermediate matrix resulting from the row transform as the input, is applied in the transform domain only to those columns predicted to contain non-zero coefficients. As an orthogonal transform, the 1D row DCT has a tendency to pack a large fraction of the signal energy into a relatively few coefficients. Therefore, the second stage of our algorithm is more effective than methods that conduct detection in the pixel domain, particularly for TBs greater than 8 × 8. Experimental results demonstrate that an HEVC encoder with our proposed two-stage algorithm can dramatically reduce unnecessary 1D row and column DCT operations compared with a standard encoder and therefore exhibits applicability to practical HEVC encoder implementation.

show abstract

Section: Related Workmentioning

confidence: 99%

“…These methods estimate the upper bound of a DCT coefficient F(u, v) as a product of a scalar α and the SAD [7,9,10]:…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A two-stage algorithm for the early detection of zero-quantized discrete cosine transform coefficients in High Efficiency Video Coding

Chen

Wang

Tian

2017

J Image Video Proc.

View full text Add to dashboard Cite

show abstract

“…In literature [1] the author used the relation between the SAD (Sum of Absolute Difference) of predicted residual block and quantization parameter (QP) to predict all-zero block. This method won't cause misjudgment, but it's efficiency is relatively low.…”

Section: Over View Of 0-value Residual Pr Ediction Based Fast Algor Imentioning

confidence: 99%

“…This method won't cause misjudgment, but it's efficiency is relatively low. The authors of literature [1] propose a new algorithm, take the SAD of each motion compensation block as the criteria to perform the all-zero block prediction. In literature [1], the author combines SAD with the DCT coefficients, deduces a more precise criteria to predict the all-zero blocks and improves the algorithm in literature [2].…”

Section: Over View Of 0-value Residual Pr Ediction Based Fast Algor Imentioning

confidence: 99%

Statistical Analysis of Residual Distribution in H.264 And A New Thought About Video Coding Algorithm

Qian¹,

Zhang²,

Sun³

2013

Proceedings of 3rd International Conference on Multimedia Technology(ICMT-13)

View full text Add to dashboard Cite

new idea about residual-based fast algorithms, and analyzes its feasibility by studying the experiment data. computation complexity is a key point that shouldn't be ignored. In the process of video coding, spatial redundancy is reduced through DCT, and temporal redundancy is reduced through predictive coding and motion compensation.Residual coding is the important content of video compression coding technology and responsible to process the data after prediction and write them into code stream and finally achieve further compression of video sequences. As the processing of residual data and related encoding make up the kernel of the code stream generating module, the property of relevant algorithms has tremendous influence on the performance of both encoder and decoder. This paper is carried out around the optimization algorithms of residual coding. propose algorithms of all-zero block prediction on the basis of study into DCT coefficient distribution models, thereby improving the encoding speed. New Ideas for Residual-based Fast Algor ithmsThe theoretical foundation of 0-value residual prediction is: when the quantization parameter is large, residuals after motion compensation in low rate 4×4 blocks tend to be very small after the DCT and quantization process, and there will be lots of all-zero coefficient blocks. Besides that there are a large number of all-zero blocks, if we can find out any other law in DCT residual distribution, new ideas for future research can be come up with.Based on the theoretical foundation of residual distribution and all-zero block prediction, we suggest that there should be distribution law in the 16 or part of the 16 quantized DCT coefficients of a 4×4 block. That is to say, among all of the 4×4 quantized DCT coefficients, there may be several types appear more frequently.Then we may consider encoding these 4×4 block types that appear in most times and transmitting their serial numbers directly, while processing the other 4×4QDCT (Quantized DCT) coefficients in conventional way. If this idea works, quite a lot of computation for processes such as entropy coding and run-length coding would be reduced. In section 5 of this paper, we'll conduct experiments based on this idea, and verify the feasibility of it by analyzing the experiment results. 1747 Pr inciple of DCT and QuantizationIn the advanced video compression standard H.264, a best mode is determined through comparison between several prediction modes, and the residuals of this best mode go through 4×4 integer discrete cosine transform to eliminate their spatial correlation. After that quantization parameters are determined according to the size of the image dynamic range, and then the DCT coefficients will be quantized according to the quantization parameter. Besides, if it is a chroma block or a intra 16×16 predicted luma block, its DC coefficients should be abstracted to form a 2×2 block (in the chroma block case) or a 4×4 block (in the intra 16×16 predicted luma block case) to go through the additional Hadamard transform, ...

show abstract

Software-Based Video Codec for Mobile Devices

Duan

Chen

et al. 2005

Embedded Software and Systems

View full text Add to dashboard Cite

Method for detecting all-zero DCT coefficients ahead of discrete cosine transformation and quantisation

Cited by 71 publications

References 1 publication

A two-stage algorithm for the early detection of zero-quantized discrete cosine transform coefficients in High Efficiency Video Coding

A two-stage algorithm for the early detection of zero-quantized discrete cosine transform coefficients in High Efficiency Video Coding

Statistical Analysis of Residual Distribution in H.264 And A New Thought About Video Coding Algorithm

Software-Based Video Codec for Mobile Devices

Contact Info

Product

Resources

About