인간 시각 특성을 이용한 웨이블릿 영상 압축 = Wavelet image compression using human visual characteristics
저자
발행사항
부산 : 부산대학교 대학원, 1998
학위논문사항
학위논문(박사)-- 부산대학교 대학원: 전자공학과 1998
발행연도
1998
작성언어
한국어
주제어
KDC
569.92 판사항(4)
발행국(도시)
부산
형태사항
vi, 98p. : 삽도 ; 26cm.
일반주기명
참고문헌수록
DOI식별코드
소장기관
In last few years discrete wavelet transform(DWT) has become well-known as powerful tool for multiresolution signal decomposition in image compression. Some promising image compression techniques with low complexity have been reported such as embedded zerotree wavelet(EZW) algorithm. the method of set partitioning in hierarchical trees(SPIHT) and stack-run coding. More complex compression methods using the joint adaptation of quantizers and filter banks have been also reported. However, these methods are designed only to improve peak signal to noise ration(PSNR) based on mean squared error(MSE). The quality of the image is judged by a human viewer. It is desirable to consider the characteristics of human visual system(HVS) in image compression. Vector quantization, discrete cosine transform, and wavelet transform based image coding incorporating human visual characteristics have been studied for a long time. In order to judge the quality of compressed images, studies of objective distortion measure rater than PSNR have been reported.
Two approaches incorporating HVS have been reported. The first approach relies on a quantization values for coefficients using subjective test. Careful subjective tests are experimentally difficult and the results obtained may vary depending on the test conditions. Therefore, image coding techniques based on subjective test are difficult to use. The second approach relies on using different quantizers formulated by mathematical models. Quantizers obtained by subjective measures can easily be used for image compression techniques, but most of them has a high computational complexity.
In this point of view, this thesis propose a simple symmetric quantizer for wavelet coefficients with low computational complexity. In this model, the scalar quantizer is obtained only by the coefficients in the lowest frequency subband. From each 2x2 block in lowest subband, the background luminance and contrast for higher frequency bands are calculated. By using these measures quantization step sized for coefficients of higher frequency subbands are determined. In the receiver, from the coefficients in lowest frequency subband which was transmitted without loss, the same quantization step sizes can be obtained.
The proposed HVS based quantizer is applied to simple adaptive scalar quantization method and EZW. Simulation results shows that the proposed method obtains almost the same performance or better than that of the previous methods in spite of lower computational complexity.
Conventionally, wavelet transform is based on circular convolution, the assumption being that the image is arranged periodically. However, because of this assumption, undesirable block boundary artifacts occur at the end of the image. The block boundary artifact is much larger when the image is compressed.
For effective wavelet-based compression, both the family and length of the wavelet coefficients must be efficiently handled. Haar filter of length two is the best for implementation, but shows the worst compression performance. There are other wavelet families that have longer and smoother filter than Haar wavelet filter. There filters provide smoother shapes upon decompression, resulting in reduction of the blocking artifacts common to the Haar transform. However, smoother wavelet functions have longer execution time. It is necessary that the criteria for selecting an appropriate wavelet transform take into account the image quality, the processing speed, and the ease of the implementation and the computational cost.
For most cases, implementations of the DWT are based upon the block processing to save memory and to permit partial area processing of interest. However, due to the block processing, the blocking artifacts occur at the block boundaries. The blocking artifacts are removed by using the hierarchical lapped transform(HLT). Additional pixels and computations are needed for using HLT. These overheads are the function of the wavelet filter length and decomposition level.
Reducing additional pixels and computational overheads is very important to design a VLSI and digital signal processor(DSP) based on compression system. In this thesis, we present and efficient wavelet transform scheme using variation of filter length. The wavelet transform scheme described in this paper uses different filter lengths in each decomposition level. In particular, by using different wavelet filter length for brightness and color components, DWT can be optimized for color image. The advantage of our approach is that the memory requirement as well as computational overheads is reduced with preserving the image quality.
서지정보 내보내기(Export)
닫기소장기관 정보
닫기권호소장정보
닫기오류접수
닫기오류 접수 확인
닫기음성서비스 신청
닫기음성서비스 신청 확인
닫기이용약관
닫기학술연구정보서비스 이용약관 (2017년 1월 1일 ~ 현재 적용)
학술연구정보서비스(이하 RISS)는 정보주체의 자유와 권리 보호를 위해 「개인정보 보호법」 및 관계 법령이 정한 바를 준수하여, 적법하게 개인정보를 처리하고 안전하게 관리하고 있습니다. 이에 「개인정보 보호법」 제30조에 따라 정보주체에게 개인정보 처리에 관한 절차 및 기준을 안내하고, 이와 관련한 고충을 신속하고 원활하게 처리할 수 있도록 하기 위하여 다음과 같이 개인정보 처리방침을 수립·공개합니다.
주요 개인정보 처리 표시(라벨링)
목 차
3년
또는 회원탈퇴시까지5년
(「전자상거래 등에서의 소비자보호에 관한3년
(「전자상거래 등에서의 소비자보호에 관한2년
이상(개인정보보호위원회 : 개인정보의 안전성 확보조치 기준)개인정보파일의 명칭 | 운영근거 / 처리목적 | 개인정보파일에 기록되는 개인정보의 항목 | 보유기간 | |
---|---|---|---|---|
학술연구정보서비스 이용자 가입정보 파일 | 한국교육학술정보원법 | 필수 | ID, 비밀번호, 성명, 생년월일, 신분(직업구분), 이메일, 소속분야, 웹진메일 수신동의 여부 | 3년 또는 탈퇴시 |
선택 | 소속기관명, 소속도서관명, 학과/부서명, 학번/직원번호, 휴대전화, 주소 |
구분 | 담당자 | 연락처 |
---|---|---|
KERIS 개인정보 보호책임자 | 정보보호본부 김태우 | - 이메일 : lsy@keris.or.kr - 전화번호 : 053-714-0439 - 팩스번호 : 053-714-0195 |
KERIS 개인정보 보호담당자 | 개인정보보호부 이상엽 | |
RISS 개인정보 보호책임자 | 대학학술본부 장금연 | - 이메일 : giltizen@keris.or.kr - 전화번호 : 053-714-0149 - 팩스번호 : 053-714-0194 |
RISS 개인정보 보호담당자 | 학술진흥부 길원진 |
자동로그아웃 안내
닫기인증오류 안내
닫기귀하께서는 휴면계정 전환 후 1년동안 회원정보 수집 및 이용에 대한
재동의를 하지 않으신 관계로 개인정보가 삭제되었습니다.
(참조 : RISS 이용약관 및 개인정보처리방침)
신규회원으로 가입하여 이용 부탁 드리며, 추가 문의는 고객센터로 연락 바랍니다.
- 기존 아이디 재사용 불가
휴면계정 안내
RISS는 [표준개인정보 보호지침]에 따라 2년을 주기로 개인정보 수집·이용에 관하여 (재)동의를 받고 있으며, (재)동의를 하지 않을 경우, 휴면계정으로 전환됩니다.
(※ 휴면계정은 원문이용 및 복사/대출 서비스를 이용할 수 없습니다.)
휴면계정으로 전환된 후 1년간 회원정보 수집·이용에 대한 재동의를 하지 않을 경우, RISS에서 자동탈퇴 및 개인정보가 삭제처리 됩니다.
고객센터 1599-3122
ARS번호+1번(회원가입 및 정보수정)