Hybrid Histogram Descriptor: A Fusion Feature Representation for Image Retrieval

Abstract

Currently, visual sensors are becoming increasingly affordable and fashionable, acceleratingly the increasing number of image data. Image retrieval has attracted increasing interest due to space exploration, industrial, and biomedical applications. Nevertheless, designing effective feature representation is acknowledged as a hard yet fundamental issue. This paper presents a fusion feature representation called a hybrid histogram descriptor (HHD) for image retrieval. The proposed descriptor comprises two histograms jointly: a perceptually uniform histogram which is extracted by exploiting the color and edge orientation information in perceptually uniform regions; and a motif co-occurrence histogram which is acquired by calculating the probability of a pair of motif patterns. To evaluate the performance, we benchmarked the proposed descriptor on RSSCN7, AID, Outex-00013, Outex-00014 and ETHZ-53 datasets. Experimental results suggest that the proposed descriptor is more effective and robust than ten recent fusion-based descriptors under the content-based image retrieval framework. The computational complexity was also analyzed to give an in-depth evaluation. Furthermore, compared with the state-of-the-art convolutional neural network (CNN)-based descriptors, the proposed descriptor also achieves comparable performance, but does not require any training process.

Local Ternary Cross Structure Pattern: A Color LBP Feature Extraction with Applications in CBIR

“…In addition, the open source implementation of the LBP feature map [9,10] can be downloaded from http://www.ee.oulu.fi/~{}gyzhao/. Referring to [6,22,23,27,40,49,50], all images in the dataset are considered as the query images to guarantee the accurateness and reproducibility. Referring to [6,22,23,27,40,49,50], the number of returned images on VisTex, KTH-2a, Face-95 and Corel-1000 is defined as 10, and the number of returned images on ETHZ and ZuBuD is defined as five because there are only five images in each class.…”

“…Finally, based on the sorting results of the similarity measure, the top similar images are returned to the user. Referring to [6,22,33,42], extended Canberra distance (ECD) is utilized in this paper, and the ECD is defined as follows:…”

“…Along with the development of imaging equipment, a larger number of images have been extensively collected from various fields [1][2][3]. Meanwhile, CBIR technology has gradually become a hot research field, due to its applications in place recognition [4], image classification [5], and remote sensing [6]. Therefore, the problem of extracting effective, robust and practical features has attracted an increasing number of researchers.…”

Special Issue on Visual Sensors

“…In this sense, Feng et al [11] proposed a texture descriptor for image retrieval designing a local parallel cross pattern in which the local binary pattern map is fused with the color map. In addition, Feng et al [12] proposed a hybrid histogram descriptor used for image retrieval. The proposed descriptor comprises two histograms jointly: a perceptual uniform histogram and a motif co-occurrence histogram including the probability of a pair of motif patterns.…”

A Local Feature Descriptor Based on Oriented Structure Maps with Guided Filtering for Multispectral Remote Sensing Image Matching

“…For example, an infrared image captures thermal radiation of the objects in a scene, while a visible image mainly records optical reflection information in a scene. Therefore, feature matching between the multispectral images is at the base of many computer vision and remote sensing applications, such as target recognition [3,4], image registration [5,6], modern military surveillance [7,8] 3D reconstruction [9,10], image fusion [11,12], and medical image processing [13,14]. Due to differences in sensors and spectra between multispectral images, there are significant nonlinear intensity variations, which mean that the grayscales and intensity distributions between multispectral images are nonlinearly correlated.…”