Taking advantage of multi-regions-based diagonal texture structure descriptor for image retrieval

Song, Wei; Zhang, Yubing; Liu, Fei; Chai, Zhilei; Ding, Feng; Xia, Qian; Park, Soon Cheol

doi:10.1016/j.eswa.2017.12.006

Cited by 28 publications

(20 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…[10] Attempted to capture high-level concepts from low-level image features under the constraint that the test image does belong to one of the classes. [11] presented their work in the area of image organization and retrieval in Image Databases using global color features and spatial color distribution of images. They suggested extending the use of image histograms to characterize the global and local color properties of an image, moreover here the artificial neural network were used.…”

Section: IImentioning

confidence: 99%

Content Based Image Retrieval based on Domain Knowledge Acquisition

2019

IJITEE

View full text Add to dashboard Cite

Enormous growth has been noted in digital technology over the years and this has led to massive growth in data mainly images and videos, hence with the volume of the data has also increased. Moreover, searching and retrieval of the accurate image from the huge dataset has become eminent as well as one of the challenging issue. In order to solve mainly two technique were used i.e. based on text and content. Moreover, content based is technique that is more efficient. In past several technique were used but they lack from the efficiency. Hence, in this paper we have proposed a methodology based named as DKA (Domain Knowledge Acquisition), DKA focuses on two main goals first is learning the similarities across the varied domain and other is learning on complex objects. The main advantage of DKA is the whether the object is whether from same domain or the other the retrieval process is very smooth. Moreover In order to evaluate the performance we have retrieved the top 5 response of query image and later it is compared with the several state-of-art technique and the existing method based on the precision parameter.

show abstract

Section: IImentioning

confidence: 99%

Content Based Image Retrieval based on Domain Knowledge Acquisition

2019

IJITEE

View full text Add to dashboard Cite

show abstract

“…Earlier years, manually managing and annotating images in large databases was time-consuming, labor-intensive, and errorprone. Thus, it is very necessary to build an efficient content-based image retrieval system [1][2][3][4][5][6][7][8][9].…”

Section: Introductionmentioning

confidence: 99%

Combination of Dominant Color Descriptor and Hu Moments in Consistent Zone for Content Based Image Retrieval

et al. 2020

View full text Add to dashboard Cite

The rapidly increasing number of digital images requires effective retrieval. Meanwhile, the dominant color descriptor has been widely used in image processing. Due to the influence of lighting and other factors, the same color in nature may have some different changes. The human eye is usually more sensitive to zones of consistent color, often identifying objects by zones of consistency. Therefore, the proposed method in this paper first applies the texton template to detect and extract the consistent zone of an image, and calculates the dominant color descriptor feature on the pixels in this consistent zone. Besides, the translation and rotation invariance of the Hu moments feature is applied to extract the shape information in the same consistent zone of the image. Finally, the combination of the dominant dolor descriptor and the Hu moments is used for content-based image retrieval. The algorithm proposed in this paper is tested on three data sets: Corel-1k, Corel-5k and Corel-10k, and the experimental results show that it is superior to the current content-based image retrieval methods.

show abstract

“…Desde la última década, se han propuesto una gran cantidad de descriptores visuales (Wang et al 2011, Tian et al, 2014, Montazar et al, 2015, Calderón et al, 2016, Meng et al, 2017, Pedronette et al, 2017, Song et al, 2018, Meng et al, 2018, Ren et al 2014, Zhu et al 2014, Jegou et al 2012y Zhang et al 2015, por citar algunos. Estos descriptores se pueden clasificar en dos categorías dependiendo del tipo de características: globales o locales.…”

Section: Introductionunclassified

“…Los descriptores globales son vectores característicos con base en colores, formas y texturas que se extraen de la imagen completa, mientras que los descriptores locales extraen características de puntos relevantes que se encuentran en cierta región de la imagen. Wang et al (2011) combinan linealmente descriptores con base en el color, la forma y la textura para recuperar imágenes relevantes con respecto a una imagen de consulta dada, mientras que Calderon et al (2016) y Song et al (2018) desarrollaron descriptores con base en textura para imágenes específicas, tales como imágenes médicas y satelitales. Montazar y Giveki (2015) propusieron un descriptor local con base en SIFT (Scale Invariant Feature Transform).…”

Section: Introductionunclassified

Redes Convolucionales Siamesas y Tripletas para la Recuperación de Imágenes Similares en Contenido

Fierro¹,

Nakano-Miyatake²,

Yanai

et al. 2019

Inf. tecnol.

View full text Add to dashboard Cite

El objetivo del trabajo presentado fue el desarrollo de un sistema de recuperación de imágenes con base en su contenido, utilizando redes convolucionales siamesas y tripletas. Se utilizaron estas arquitecturas múltiples para generar descriptores visuales, extrayendo información semántica de dos imágenes (siamesa) o tres imágenes (tripleta) a la vez. Posteriormente, se realizó un aprendizaje de similitud, codificando la distancia de estas siamesas o tripletas de descriptores visuales, cuyo almacenamiento no es necesario. Los resultados muestran que los esquemas con base en redes convolucionales extraen mayor cantidad de información semántica. Las arquitecturas múltiples, aparte de extraer información semántica, mejoran la tasa de recuperación de imágenes. Se concluye que las arquitecturas múltiples solucionan los tres retos más importantes de estos sistemas, como lo son la brecha semántica, el aprendizaje de similitud y el espacio de almacenamiento, los cuales no habían sido resueltos en trabajos anteriores.

show abstract

Taking advantage of multi-regions-based diagonal texture structure descriptor for image retrieval

Cited by 28 publications

References 39 publications

Content Based Image Retrieval based on Domain Knowledge Acquisition

Content Based Image Retrieval based on Domain Knowledge Acquisition

Combination of Dominant Color Descriptor and Hu Moments in Consistent Zone for Content Based Image Retrieval

Redes Convolucionales Siamesas y Tripletas para la Recuperación de Imágenes Similares en Contenido

Contact Info

Product

Resources

About