### Citations

1995 |
Pattern Recognition with Fuzzy Objective Function Algorithm
- Bezdek
- 1981
(Show Context)
Citation Context ...image I = {R1, R2,…, Rn) where Rk is the k region of image I. • I = ∪ Ri where i = 1, 2, …, n. • Ri ∩ Rj = φ for all i≠j.s253 In this research, the well known statistical clustering algorithm, K-mean,=-=[3,4,5]-=- is used to segment images into disjoint regions. Then regions that represent objects or meaningful parts are identified by the statistical goodness of fit test. The objective of the region identifica... |

1602 | Color Indexing
- Swain, Ballard
- 1991
(Show Context)
Citation Context ...n the region Histogram indexing technique (HIST) is the most common visual feature that has been used by many image retrieval systems such as FINDIT system developed by M. J. Swain and his colleagues =-=[7]-=- and QUBIC system from IBM Corporation. The histogram computed from each region and retrieval is done by comparing the histogram of the query’s regions with histograms of all regions in the image coll... |

364 |
Pattern recognition principles
- Tou, Gonzales
- 1974
(Show Context)
Citation Context ...image I = {R1, R2,…, Rn) where Rk is the k region of image I. • I = ∪ Ri where i = 1, 2, …, n. • Ri ∩ Rj = φ for all i≠j.s253 In this research, the well known statistical clustering algorithm, K-mean,=-=[3,4,5]-=- is used to segment images into disjoint regions. Then regions that represent objects or meaningful parts are identified by the statistical goodness of fit test. The objective of the region identifica... |

326 |
Texture classification and segmentation using wavelet frames
- Unser
- 1995
(Show Context)
Citation Context ...equency bands of onelevel Daubechies wavelet transform of the three-color components. The wavelet features capture the spatial color variations and texture properties of visual data effectively Unser =-=[6]-=-. Figure (4) and (5) show segmentation results for some images with non-object (texture) and object regions. 2.3 Classification of Object and Non-Object Regions The classification of regions into obje... |

57 | Automated binary texture feature sets for image retrieval
- Smith, Chang
- 1996
(Show Context)
Citation Context ...in image compression. Several wavelet-based indexing techniques have been proposed in the literature of CBIR. Histogram of wavelet coefficients of different bands has been proposed by Smith and Chang =-=[9]-=-. The discrete cosine transform of the wavelet lowestfrequency subband (WDCT) has been proposed by Abbadi [10]. The magnitude of the wavelet coefficients (WMGT) has been proposed by Wang et al. [11]. ... |

54 |
Color matching for image retrieval
- Mehtre, Kankanhalli, et al.
- 1994
(Show Context)
Citation Context ...retrieved that have least differences of regions histograms. It has been shown that the histogram provides as260 good indexing and retrieval performance while being inexpensive in term of computation =-=[8]-=-. The histogram-based indexing is invariant to translation and rotation of regions. The histogram provides a good indexing and retrieval performance for regions of large objects but it fails to provid... |

2 |
Document N390R, New Work Item: JPEG 2000 image coding system
- JTC1SC29WG1
- 1997
(Show Context)
Citation Context ...d in which images are indexed by their own visual features that are extracted automatically. CBIR have become important with recent advances in image and video compression standards such as JPEG-2000 =-=[1]-=- and MPEG-4 [2]. They have wide applications in several areas such as multimedia information systems and digital libraries for education, medical imaging, remote sensing, movie industry, and video on ... |

2 |
Content-based indexing and searching in image databases. Doctorial dissertation
- Abbadi
- 2000
(Show Context)
Citation Context ... Histogram of wavelet coefficients of different bands has been proposed by Smith and Chang [9]. The discrete cosine transform of the wavelet lowestfrequency subband (WDCT) has been proposed by Abbadi =-=[10]-=-. The magnitude of the wavelet coefficients (WMGT) has been proposed by Wang et al. [11]. In the Wavelet Histogram Technique (WHST) proposed by Smith and Chang [9], the image is first decomposed M tim... |

2 |
et al. Waveletbased image indexing techniques with partial sketch retrieval capability
- Wang, Wiederhold, et al.
- 1997
(Show Context)
Citation Context ...ng [9]. The discrete cosine transform of the wavelet lowestfrequency subband (WDCT) has been proposed by Abbadi [10]. The magnitude of the wavelet coefficients (WMGT) has been proposed by Wang et al. =-=[11]-=-. In the Wavelet Histogram Technique (WHST) proposed by Smith and Chang [9], the image is first decomposed M times to wavelet bands that form a pyramid of M levels. Each level consist of three high-pa... |

1 |
Region clustering and identification for advanced Content-Based Image retrieval systems" The George Washington University 2002
- Ameen
(Show Context)
Citation Context ...e regions that are result from the region identification process described in section 2 can be indexed by mathematical measures extracted from a window-constructed around each region by padding zeros =-=[12]-=-. The mathematical measures are also called features such as color, texture, and shape represent different properties. In this section, four general domain image-indexing techniques are presented. The... |