Remotely sensed image retrieval based on region-level semantic mining

Liu, Tingting; Zhang, Liangpei; Li, Pingxiang; Lin, Hui

doi:10.1186/1687-5281-2012-4

Cited by 30 publications

(10 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More recently, Samal et al [17] investigated autocorrelation measures and fractal attributes computed from a hierarchical image representation, and Gleason et al [18] combined multispectral histograms with LBP and LEP. In addition, Liu et al [19] studied a region-based approach where a combination of color and texture features are computed on image regions after segmentation.…”

Section: A Content-based Remote Sensing Image Retrievalmentioning

confidence: 99%

Remote Sensing Image Retrieval With Global Morphological Texture Descriptors

Aptoula

2014

IEEE Trans. Geosci. Remote Sensing

170

127

View full text Add to dashboard Cite

In this paper, we present the results of applying global morphological texture descriptors to the problem of content-based remote sensing image retrieval. Specifically, we explore the potential of recently developed multiscale texture descriptors, namely, the circular covariance histogram and the rotation-invariant point triplets. Moreover, we introduce a couple of new descriptors, exploiting the Fourier power spectrum of the quasi-flat-zone-based scale space of their input. The descriptors are evaluated with the UC Merced Land Use-Land Cover data set, which has been only recently made public. The proposed approach is shown to outperform the best known retrieval scores, despite its shorter feature vector length, thus asserting the practical interest of global content descriptors as well as of mathematical morphology in this context. Index Terms-Content-based image retrieval (CBIR), mathematical morphology (MM), remote sensing, texture description.

show abstract

Section: A Content-based Remote Sensing Image Retrievalmentioning

confidence: 99%

Remote Sensing Image Retrieval With Global Morphological Texture Descriptors

Aptoula

2014

IEEE Trans. Geosci. Remote Sensing

170

127

View full text Add to dashboard Cite

show abstract

“…In other words, the semantic contents of an image cannot be well revealed by these features. To alleviate this issue, Liu et al [17] proposed a region-level semantic mining approach for image presentation and constructed a uniform region-based depiction for each image by segmenting the images by region. Then, the semantic features were extracted using a probabilistic method, which had good retrieval precision and recall.…”

Section: Background and Related Studiesmentioning

confidence: 99%

High-Resolution Remote Sensing Image Retrieval Based on CNNs from a Dimensional Perspective

Xiao

Yang

et al. 2017

Remote Sensing

View full text Add to dashboard Cite

Because of recent advances in Convolutional Neural Networks (CNNs), traditional CNNs have been employed to extract thousands of codes as feature representations for image retrieval. In this paper, we propose that more powerful features for high-resolution remote sensing image representations can be learned using only several tens of codes; this approach can improve the retrieval accuracy and decrease the time and storage requirements. To accomplish this goal, we first investigate the learning of a series of features with different dimensions using a few tens to thousands of codes via our improved CNN frameworks. Then, a Principal Component Analysis (PCA) is introduced to compress the high-dimensional remote sensing image feature codes learned by traditional CNNs. Comprehensive comparisons are conducted to evaluate the retrieval performance based on feature codes of different dimensions learned by the improved CNNs as well as the PCA compression. To further demonstrate the powerful ability of the low-dimensional feature representation learned by the improved CNN frameworks, a Feature Weighted Map (FWM), which can perform feature visualization and provides a better understanding of the nature of Deep Convolutional Neural Networks (DCNNs) frameworks, is explored. All the CNN models are trained from scratch using a large-scale and high-resolution remote sensing image archive, which will be published and made available to the public. The experimental results show that our method outperforms state-of-the-art CNN frameworks in terms of accuracy and storage.

show abstract

“…Samal et al 6 used integrated signature and a quadtree-based idea in a retrieval engine which provided an efficient approach to matching remote sensing imagery. Liu et al 7 described a regionlevel satellite image retrieval system in which a region-level semantic mining approach was proposed to narrow the "semantic gap".…”

Section: Introductionmentioning

confidence: 99%

Rotation-invariant image retrieval using hidden Markov tree for remote sensing data

Miao

Zhao

2014

SPIE Proceedings

View full text Add to dashboard Cite

The rapid increase in quantity of available remote sensing data brought an urgent need for intelligent retrieval techniques for remote sensing images. As one of the basic visual characteristics and important information sources of remote sensing images, texture is widely used in the scheme of remote sensing image retrieval. Since many images or regions with identical texture features usually show the diversity of direction, the consideration of rotation-invariance in the description of texture features is of significance both theoretically and practically. To address these issues, we develop a rotation-invariant image retrieval method based on the texture features of remote sensing images. We use the steerable pyramid transform to get the multi-scale and multi-orientation representation of texture images. Then we employ the hidden Markov tree (HMT) model, which provides a good tool to describe texture feature, to capture the dependencies across scales and orientations, by which the statistical properties of the transform domain coefficients can be obtained. Utilizing the inherent tree structure of the HMT and its fast training and likelihood computation algorithms, we can extract the rotation-invariant features of texture images. Similarity between the query image and each candidate image in the database can be measured by computing the Kullback-Leibler distance between the corresponding models. We evaluate the retrieval effectiveness of the algorithm with Brodatz texture database and remote sensing images. The experimental results show that this method has satisfactory performance in image retrieval and less sensitivity to texture rotation.

show abstract

Remotely sensed image retrieval based on region-level semantic mining

Cited by 30 publications

References 24 publications

Remote Sensing Image Retrieval With Global Morphological Texture Descriptors

Remote Sensing Image Retrieval With Global Morphological Texture Descriptors

High-Resolution Remote Sensing Image Retrieval Based on CNNs from a Dimensional Perspective

Rotation-invariant image retrieval using hidden Markov tree for remote sensing data

Contact Info

Product

Resources

About