Simplified realization of normalized integer WHT for multiplierless integer DCT

Taizo Suzuki, Masaaki Ikehara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Walsh-Hadamard transform (WHT) based multiplierless integer discrete cosine transform (IntDCT) has structural regularity even in short word length lifting coefficients. It, however, cannot apply to image coding without the quantization part because WHT was implemented by only 1 adder operations without the normalization scaling factors. Although we have already presented a normalized integer WHT (IntWHT) as its solution, it also has many adder operations. In this paper, using a two-dimensional (2-D) separable transform of one-dimensional (1-D) normalized WHT is applied to each lifting coefficient, we present a more simplified realization of normalized IntWHT with structural regularity for short word length lifting coefficients. Finally, in lossless-to-lossy image coding, IntDCT based on the proposed IntWHT is validated.

Original languageEnglish
Title of host publication2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings
Pages472-476
Number of pages5
DOIs
Publication statusPublished - 2011
Event2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Sedona, AZ, United States
Duration: 2011 Jan 42011 Jan 7

Other

Other2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011
CountryUnited States
CitySedona, AZ
Period11/1/411/1/7

Fingerprint

Walsh transforms
Hadamard transforms
regularity
coding
Discrete cosine transforms
Adders
normalization
Image coding
scaling

Keywords

  • Integer discrete cosine transform (IntDCT)
  • integer Walsh-Hadamard transform (IntWHT)
  • lossless-to-lossy image coding

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Signal Processing
  • Education

Cite this

Suzuki, T., & Ikehara, M. (2011). Simplified realization of normalized integer WHT for multiplierless integer DCT. In 2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings (pp. 472-476). [5739260] https://doi.org/10.1109/DSP-SPE.2011.5739260

Simplified realization of normalized integer WHT for multiplierless integer DCT. / Suzuki, Taizo; Ikehara, Masaaki.

2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings. 2011. p. 472-476 5739260.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Suzuki, T & Ikehara, M 2011, Simplified realization of normalized integer WHT for multiplierless integer DCT. in 2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings., 5739260, pp. 472-476, 2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011, Sedona, AZ, United States, 11/1/4. https://doi.org/10.1109/DSP-SPE.2011.5739260
Suzuki T, Ikehara M. Simplified realization of normalized integer WHT for multiplierless integer DCT. In 2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings. 2011. p. 472-476. 5739260 https://doi.org/10.1109/DSP-SPE.2011.5739260
Suzuki, Taizo ; Ikehara, Masaaki. / Simplified realization of normalized integer WHT for multiplierless integer DCT. 2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings. 2011. pp. 472-476
@inproceedings{186c272a52f2492a8514f41db4dd8db6,
title = "Simplified realization of normalized integer WHT for multiplierless integer DCT",
abstract = "Walsh-Hadamard transform (WHT) based multiplierless integer discrete cosine transform (IntDCT) has structural regularity even in short word length lifting coefficients. It, however, cannot apply to image coding without the quantization part because WHT was implemented by only 1 adder operations without the normalization scaling factors. Although we have already presented a normalized integer WHT (IntWHT) as its solution, it also has many adder operations. In this paper, using a two-dimensional (2-D) separable transform of one-dimensional (1-D) normalized WHT is applied to each lifting coefficient, we present a more simplified realization of normalized IntWHT with structural regularity for short word length lifting coefficients. Finally, in lossless-to-lossy image coding, IntDCT based on the proposed IntWHT is validated.",
keywords = "Integer discrete cosine transform (IntDCT), integer Walsh-Hadamard transform (IntWHT), lossless-to-lossy image coding",
author = "Taizo Suzuki and Masaaki Ikehara",
year = "2011",
doi = "10.1109/DSP-SPE.2011.5739260",
language = "English",
isbn = "9781612842271",
pages = "472--476",
booktitle = "2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings",

}

TY - GEN

T1 - Simplified realization of normalized integer WHT for multiplierless integer DCT

AU - Suzuki, Taizo

AU - Ikehara, Masaaki

PY - 2011

Y1 - 2011

N2 - Walsh-Hadamard transform (WHT) based multiplierless integer discrete cosine transform (IntDCT) has structural regularity even in short word length lifting coefficients. It, however, cannot apply to image coding without the quantization part because WHT was implemented by only 1 adder operations without the normalization scaling factors. Although we have already presented a normalized integer WHT (IntWHT) as its solution, it also has many adder operations. In this paper, using a two-dimensional (2-D) separable transform of one-dimensional (1-D) normalized WHT is applied to each lifting coefficient, we present a more simplified realization of normalized IntWHT with structural regularity for short word length lifting coefficients. Finally, in lossless-to-lossy image coding, IntDCT based on the proposed IntWHT is validated.

AB - Walsh-Hadamard transform (WHT) based multiplierless integer discrete cosine transform (IntDCT) has structural regularity even in short word length lifting coefficients. It, however, cannot apply to image coding without the quantization part because WHT was implemented by only 1 adder operations without the normalization scaling factors. Although we have already presented a normalized integer WHT (IntWHT) as its solution, it also has many adder operations. In this paper, using a two-dimensional (2-D) separable transform of one-dimensional (1-D) normalized WHT is applied to each lifting coefficient, we present a more simplified realization of normalized IntWHT with structural regularity for short word length lifting coefficients. Finally, in lossless-to-lossy image coding, IntDCT based on the proposed IntWHT is validated.

KW - Integer discrete cosine transform (IntDCT)

KW - integer Walsh-Hadamard transform (IntWHT)

KW - lossless-to-lossy image coding

UR - http://www.scopus.com/inward/record.url?scp=79954477512&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79954477512&partnerID=8YFLogxK

U2 - 10.1109/DSP-SPE.2011.5739260

DO - 10.1109/DSP-SPE.2011.5739260

M3 - Conference contribution

AN - SCOPUS:79954477512

SN - 9781612842271

SP - 472

EP - 476

BT - 2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings

ER -