Ro, Yong-Man
(Information and Communications University)
,
Kim, Mun-Churl
(Broadcasting Media Technology Department, ETRI)
,
Kang, Ho-Kyung
(Information and Communications University)
,
Manjunath, B.S.
(University of California)
,
Kim, Jin-Woong
(Broadcasting Media Technology Department, ETRI)
MPEG-7 standardization work has started with the aims of providing fundamental tools for describing multimedia contents. MPEG-7 defines the syntax and semantics of descriptors and description schemes so that they may be used as fundamental tools for multimedia content description. In this paper, we ...
MPEG-7 standardization work has started with the aims of providing fundamental tools for describing multimedia contents. MPEG-7 defines the syntax and semantics of descriptors and description schemes so that they may be used as fundamental tools for multimedia content description. In this paper, we introduce a texture based image description and retrieval method, which is adopted as the homogeneous texture descriptor in the visual part of the MPEG-7 final committee draft. The current MPEG-7 homogeneous texture descriptor consists of the mean, the standard deviation value of an image, energy, and energy deviation values of Fourier transform of the image. These are extracted from partitioned frequency channels based on the human visual system (HVS). For reliable extraction of the texture descriptor, Radon transformation is employed. This is suitable for HVS behavior. We also introduce various matching methods; for example, intensity-invariant, rotation-invariant and/or scale-invariant matching. This technique retrieves relevant texture images when the user gives a querying texture image. In order to show the promising performance of the texture descriptor, we take the experimental results with the MPEG-7 test sets. Experimental results show that the MPEG-7 texture descriptor gives an efficient and effective retrieval rate. Furthermore, it gives fast feature extraction time for constructing the texture descriptor.
MPEG-7 standardization work has started with the aims of providing fundamental tools for describing multimedia contents. MPEG-7 defines the syntax and semantics of descriptors and description schemes so that they may be used as fundamental tools for multimedia content description. In this paper, we introduce a texture based image description and retrieval method, which is adopted as the homogeneous texture descriptor in the visual part of the MPEG-7 final committee draft. The current MPEG-7 homogeneous texture descriptor consists of the mean, the standard deviation value of an image, energy, and energy deviation values of Fourier transform of the image. These are extracted from partitioned frequency channels based on the human visual system (HVS). For reliable extraction of the texture descriptor, Radon transformation is employed. This is suitable for HVS behavior. We also introduce various matching methods; for example, intensity-invariant, rotation-invariant and/or scale-invariant matching. This technique retrieves relevant texture images when the user gives a querying texture image. In order to show the promising performance of the texture descriptor, we take the experimental results with the MPEG-7 test sets. Experimental results show that the MPEG-7 texture descriptor gives an efficient and effective retrieval rate. Furthermore, it gives fast feature extraction time for constructing the texture descriptor.
* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.
문제 정의
In this paper, we present technical details of the MPEG-7 homogeneous texture descriptor and its feature extraction method. In addition to the feature extraction method, similarity measuring criteria are presented for rotation-, scale- and inten sity-invariant matchings[15]-[18].
제안 방법
In this paper, we introduce a texture-based image description and retrieval method which we proposed and adopted as the Homogeneous texture descriptor in the Visual pait of the MPEG-7 FCD. Our proposal was adopted.
The 70 images were rotated with 0, 8, 25, 55, 107, 131, and 174 degrees. Then, the rotated images are scaled with 90%, 80%, 70%, 60% and 50%. Finally, the T7 data set is constructed by taking 128x128 size image at aibitraiy position from both rotated and scaled images.
To verify the performance of the MPEG-7 texture descriptor, experiments were performed with test data sets of the homoge neous texture descriptor. These are T1, T2, T3, T4, T5, T6, and T7 data sets.
대상 데이터
32 % of AVRR was the result for the enhancement layer. 62 components of the descriptor were used. For the T1 data set, half of the description size could be saved with only about 1% loss of AVRR.
T1 data set contains texture pattern images which have been used popularly as a test image set for the texture experiments in many literatures. It consists of 1856 images with matrix size of 128×128. 1856 images are made from 116 Brodatz images with matrix size of 512×512.
T2 data set consists of real patterns taken from outdoor and indoor scenes. It consists of 832 images with size of 128×128. Like the T1 data set, 832 images in T2 data set are made from 52 images with matrix size of 512x512 such that an image of matrix size of 512×512 is divided into 16 non-overlapped parti tions.
39 % of AVRR was obtained with the base layer in the T1 data set. Only 32 components of the descriptor were used. 77.
The MPEG-7 test data sets for the texture descriptor have 7 different kinds of test data sets, which are T1, T2, T3, T4, T5, T6, and T7 data sets. The following subsections explain MPEG-7 test data sets used in the core experiments of the ho mogeneous texture descriptor in detail.
이론/모형
It provided higher retrieval accuracy for the testing data sets. Therefore, the homogeneous texture descriptor de scribed in this paper was selected as the normative MPEG-7 homogeneous texture descriptor in the Visual part of the MPEG-7 final committee draft[11]-[14].
참고문헌 (7)
Faugeras, Olivier D., Pratt, William K..
Decorrelation Methods of Texture Feature Extraction.
IEEE transactions on pattern analysis and machine intelligence,
vol.pami2,
no.4,
323-332.
Manjunath, B.S., Ma, W.Y..
Texture features for browsing and retrieval of image data.
IEEE transactions on pattern analysis and machine intelligence,
vol.18,
no.8,
837-842.
Kim, Young-Sum, Kim, Yong-Sung, Kim, Whoi-Yul, Kim, Myung-Joon.
Development of Content-Based Trademark Retrieval System on the World Wide Web.
ETRI journal,
vol.21,
no.1,
40-54.
Daugman, J.G..
High confidence visual recognition of persons by a test of statistical independence.
IEEE transactions on pattern analysis and machine intelligence,
vol.15,
no.11,
1148-1161.
Chang, T., Kuo, C.-C.J..
Texture analysis and classification with tree-structured wavelet transform.
IEEE transactions on image processing : a publication of the IEEE Signal Processing Society,
vol.2,
no.4,
429-441.
AI-Helper
※ AI-Helper는 부적절한 답변을 할 수 있습니다.