Higher Retrieval Accuracy Research Articles

Analysis of Heterogeneous Data Integration and Document Clustering method in Digital Library

This paper proposes a hierarchical organization framework model of heterogeneous data integration and resource utilization of digital library. The model is divided into metadata layer, ontology layer, link data layer and application layer from top to bottom. Aiming at the drawbacks of the traditional document clustering methods, an improved LDA method is proposed to cluster the shared subject documents of the same category. The results show that the retrieval time of the proposed linked data organization framework model of digital library is much smaller than that of the traditional model, and the search response time is reduced by 56.3%. The establishment of the linked data organization framework model of digital library can greatly increase the interconnection of distributed heterogeneous information and make the data source more easily crawled by search engines. TC-LDA algorithm has the highest retrieval accuracy and the best stability. The time consumption of the five algorithms is almost the same, thus it hard to get the algorithm with advantage of time-consuming. The improved LDA text subject clustering algorithm has more advantages than the traditional text clustering method, which is obtained from the analysis of clustering accuracy and time consuming.

The AOTF-based NO<sub>2</sub> camera

Abstract. The abundance of NO2 in the boundary layer relates to air quality and pollution source monitoring. Observing the spatiotemporal distribution of NO2 above well-delimited (flue gas stacks, volcanoes, ships) or more extended sources (cities) allows for applications such as monitoring emission fluxes or studying the plume dynamic chemistry and its transport. So far, most attempts to map the NO2 field from the ground have been made with visible-light scanning grating spectrometers. Benefiting from a high retrieval accuracy, they only achieve a relatively low spatiotemporal resolution that hampers the detection of dynamic features. We present a new type of passive remote sensing instrument aiming at the measurement of the 2-D distributions of NO2 slant column densities (SCDs) with a high spatiotemporal resolution. The measurement principle has strong similarities with the popular filter-based SO2 camera as it relies on spectral images taken at wavelengths where the molecule absorption cross section is different. Contrary to the SO2 camera, the spectral selection is performed by an acousto-optical tunable filter (AOTF) capable of resolving the target molecule's spectral features. The NO2 camera capabilities are demonstrated by imaging the NO2 abundance in the plume of a coal-fired power plant. During this experiment, the 2-D distribution of the NO2 SCD was retrieved with a temporal resolution of 3 min and a spatial sampling of 50 cm (over a 250 × 250 m2 area). The detection limit was close to 5 × 1016 molecules cm−2, with a maximum detected SCD of 4 × 1017 molecules cm−2. Illustrating the added value of the NO2 camera measurements, the data reveal the dynamics of the NO to NO2 conversion in the early plume with an unprecedent resolution: from its release in the air, and for 100 m upwards, the observed NO2 plume concentration increased at a rate of 0.75–1.25 g s−1. In joint campaigns with SO2 cameras, the NO2 camera could also help in removing the bias introduced by the NO2 interference with the SO2 spectrum.

Validation of the generalized single-channel algorithm using Landsat 8 imagery and SURFRAD ground measurements

ABSTRACTLand surface temperature (LST) is a key parameter in land surface physics and depends on energy and water exchange processes with the atmosphere. The generalized single-channel (GSC) algorithm is an operational approach for retrieving LST from satellite sensors with only one thermal infrared channel. Until now, validations of the GSC algorithm using in situ data were limited. In this study, Landsat 8 imagery and SURFace RADiation budget network (SURFRAD) ground observations were employed to test the accuracy of the GSC algorithm. The validation results show that the GSC algorithm can obtain high LST retrieval accuracy when the total atmospheric water vapour content is between 0.5 and 3 g cm−2, with coefficient of determination (R2) between the retrieved LST and the ground LST greater than 0.97 and mean absolute error (MAE) and root mean square error (RMSE) values of 1.57 K and 1.96 K, respectively.

Boosting local texture descriptors with Log-Gabor filters response for improved image retrieval

In the recent past, many local texture descriptors have been proposed for the image retrieval task. In order to improve the image retrieval accuracy, quite a few of these descriptors have been implemented on Gabor filter response. However, the response of Log-Gabor filters has been proved to be better than Gabor filters with respect to their discrimination ability. In this paper, we present a framework for image retrieval that applies various local texture descriptors on Log-Gabor filters response. To evaluate the retrieval performance of the proposed framework, experiments have been conducted on standard Wang, VisTex and OT-Scene databases. Consistent improvement in the image retrieval accuracy demonstrates the effectiveness of this framework. Further, the experimental results show that the use of proposed framework with low-dimension texture descriptors such as Orthogonal Combination of Local Binary Pattern makes them a better choice over Local Binary Pattern and its high-dimensional variants when higher retrieval accuracy, small feature vector size and ease of computation is desired.

Enhancing CBIR Performance Using Evolutionary Algorithm-Assisted Significant Feature Selection: A Filter Approach

Efficient searching becomes essential for large image archive, with more and more digital images available on Internet. Consequently, content-based image retrieval (CBIR) has drawn widespread research attentiveness in the last decade in the field of image processing, pattern recognition and computer vision. CBIR approach boils down to two core problems: feature extraction followed by feature matching. Feature selection is a process that selects pertinent features as a subset of original extracted features. This paper presents six filter approaches for significant features’ selection: decision tree, relief, genetic algorithm (GA) with correlation based feature selection (CFS) as fitness function, particle swarm optimisation (PSO) with CFS as fitness function, exhaustive search and forward selection with CFS as attribute subset evaluator. Work is carried out on publicly available Corel data set images. Experimental results prove that the features selected from PSO and GA with CFS enhances CBIR performance both in terms of higher retrieval accuracy and reduced computational time.

지역 색차 기반의 히스토그램 정교화에 의한 영상 검색

Since digital images and videos are rapidly increasing in the internet with the spread of mobile computers and smartphones, research on image retrieval has gained tremendous momentum. Color, shape, and texture are major features used in image retrieval. Especially, color information has been widely used in image retrieval, because it is robust in translation, rotation, and a small change of camera view. This paper proposes a new method for histogram refinement based on local color difference. Firstly, the proposed method converts a RGB color image into a HSV color image. Secondly, it reduces the size of color space from 256 to 32. It classifies pixels in the 32-color image into three groups according to the color difference between a central pixel and its neighbors in a 3x3 local region. Finally, it makes a color difference vector(CDV) representing three refined color histograms, then image retrieval is performed by the CDV matching. The experimental results using public image database show that the proposed method has higher retrieval accuracy than other conventional ones. They also show that the proposed method can be effectively applied to search low resolution images such as thumbnail images.

Support Vector Machine Modeling Using Particle Swarm Optimization Approach for the Retrieval of Atmospheric Ammonia Concentrations

This study was performed in order to improve the estimation accuracy of atmospheric ammonia (NH3) concentration levels in the Greater Houston area during extended sampling periods. The approach is based on selecting the appropriate penalty coefficient C and kernel parameter σ 2. These parameters directly influence the regression accuracy of the support vector machine (SVM) model. In this paper, two artificial intelligence techniques, particle swarm optimization (PSO) and a genetic algorithm (GA), were used to optimize the SVM model parameters. Data regarding meteorological variables (e.g., ambient temperature and wind direction) and the NH3 concentration levels were employed to develop our two models. The simulation results indicate that both PSO-SVM and GA-SVM methods are effective tools to model the NH3 concentration levels and can yield good prediction performance based on statistical evaluation criteria. PSO-SVM provides higher retrieval accuracy and faster running speed than GA-SVM. In addition, we used the PSO-SVM technique to estimate 17 drop-off NH3 concentration values. We obtained forecasting results with good fitting characteristics to a measured curve. This proved that PSO-SVM is an effective method for estimating unavailable NH3 concentration data at 3, 4, 5, and 6 parts per billion (ppb), respectively. A 4-ppb NH3 concentration had the optimum prediction performance of the simulation results. These results showed that the selection of the set-point values is a significant factor in compensating for the atmospheric NH3 dropout data with the PSO-SVM method. This modeling approach will be useful in the continuous assessment of NH3 sensor discrete data sources.

Forest Data Collection Using Terrestrial Image-Based Point Clouds From a Handheld Camera Compared to Terrestrial and Personal Laser Scanning

Stereo images have long been the main practical data source for the high-accuracy retrieval of 3-D information over large areas. However, stereoscopy has been surpassed by laser scanning (LS) techniques in recent years, particularly in forested areas, because the reflection of laser points from object surfaces directly provides 3-D geometric features and because the laser beam has good penetration capacity through forest canopies. In the last few years, image-based point clouds have become a more widely available data source because of advances in matching algorithms and computer hardware. This paper explores the possibility of using consumer cameras for forest field data collection and presents an application of terrestrial image-based point clouds derived from a handheld camera to forest plot inventories. In the experiment, the sample forest plot was photographed in a stop-and-go mode using different routes and camera settings. Five data sets were generated from photographs taken in the field, representing different photographic conditions. The stem detection accuracy ranged between 60% and 84%, and the root-mean-square errors of the estimated diameters at breast height were between 2.98 and 6.79 cm. The performance of image-based point clouds in forest data collection was compared with that of point clouds derived from two LS techniques, i.e., terrestrial LS (the professional level) and personal LS (an emerging technology). The study indicates that the construction of image-based point clouds of forest field data requires only low-cost, low-weight, and easy-to-use equipment and automated data processing. Photographic measurement is easy and relatively fast. The accuracy of tree attribute estimates is close to an acceptable level for forest field inventory but is lower than that achieved with the tested LS techniques.

Permutation search methods are efficient, yet faster search is possible

We survey permutation-based methods for approximate k -nearest neighbor search. In these methods, every data point is represented by a ranked list of pivots sorted by the distance to this point. Such ranked lists are called permutations. The underpinning assumption is that, for both metric and non-metric spaces, the distance between permutations is a good proxy for the distance between original points. Thus, it should be possible to efficiently retrieve most true nearest neighbors by examining only a tiny subset of data points whose permutations are similar to the permutation of a query. We further test this assumption by carrying out an extensive experimental evaluation where permutation methods are pitted against state-of-the art benchmarks (the multi-probe LSH, the VP-tree, and proximity-graph based retrieval) on a variety of realistically large data set from the image and textual domain. The focus is on the high-accuracy retrieval methods for generic spaces. Additionally, we assume that both data and indices are stored in main memory. We find permutation methods to be reasonably efficient and describe a setup where these methods are most useful. To ease reproducibility, we make our software and data sets publicly available.

Improving Microblog Retrieval from Exterior Corpus by Automatically Constructing Microblogging Corpus

A large-scale training corpus consisting of microblogs belonging to a desired category is important for high-accuracy microblog retrieval. Obtaining such a large-scale microblgging corpus manually is very time and labor-consuming. Therefore, some models for the automatic retrieval of microblogs froman exterior corpus have been proposed. However, these approaches may fail in considering microblog-specific features. To alleviate this issue, we propose a methodology that constructs a simulated microblogging corpus rather than directly building a model from the exterior corpus. The performance of our model is better since the microblog-special knowledge of the microblogging corpus is used in the end by the retrieval model. Experimental results on real-world microblogs demonstrate the superiority of our technique compared to the previous approaches.

Locality Preserving Vector and Image-Specific Topic Model for Visual Recognition

Nowadays, evolution of mobile devices make demand for searching information increasing expressively. Many applications have been developed for recognition tasks. In this paper, we present a new and efficient visual search system for finding similar images on the large database. We first propose a compact, discriminative image representation called Locality Preserving Vector which can explicitly exploit neighborhood structure of data and attains high retrieval accuracy in the low-dimensional space. We then integrate topic modeling into visual search system for extracting topic related and image-specific information. These information enables images which likely contain the same objects to be ranked with higher similarity. The experiments show that our approach provides competitive accuracy with very low memory cost.

小波域Copula 多维模型纹理检索

In this paper, an effective texture retrieval method using copula multidimensional model on wavelet domain is proposed. For make up for the shortcomings of the single statistical model on wavelet subband, a treedependence structure on wavelet domain is constructed, and copula multidimensional distribution is implemented on the tree-dependence structure of wavelet domain. Tree-dependence structure can capture both the dependences of inter-scale and neighbor-dependence structures, and it has fewer dimension and fewer number of copula models comparing to the neighbor-dependence structure. Because of the complexity of the copula multidimensional model, it is difficult to deduce the Kullback-Leibler distance (KLD) of copula model. This paper puts forward a kind of retrieval method based on KLD of copula multidimensional model, which is consisted of two components: The KLD of marginal distributions and the KLD of copula function. The experimental results on VisTex and Brodatz databases show that the proposed retrieval method has low computational complexity and high retrieval accuracy, and it is more effective than the state-of-the-art copula methods on wavelet domain. The proposed texture retrieval method can be extended to other wavelet domains such as complex wavelet and directional wavelet domains.

Retrieving self-vocalized information: An event-related potential (ERP) study on the effect of retrieval orientation

Retrieval orientation refers to a pre-retrieval process and conceptualizes the specific form of processing that is applied to a retrieval cue. In the current event-related potential (ERP) study, we sought to find evidence for an involvement of the auditory cortex when subjects attempt to retrieve vocalized information, and hypothesized that adopting retrieval orientation would be beneficial for retrieval accuracy. During study, participants saw object words that they subsequently vocalized or visually imagined. At test, participants had to identify object names of one study condition as targets and to reject object names of the second condition together with new items. Target category switched after half of the test trials. Behaviorally, participants responded less accurately and more slowly to targets of the vocalize condition than to targets of the imagine condition. ERPs to new items varied at a single left electrode (T7) between 500 and 800ms, indicating a moderate retrieval orientation effect in the subject group as a whole. However, whereas the effect was strongly pronounced in participants with high retrieval accuracy, it was absent in participants with low retrieval accuracy. A current source density (CSD) mapping of the retrieval orientation effect indicated a source over left temporal regions. Independently from retrieval accuracy, the ERP retrieval orientation effect was surprisingly also modulated by test order. Findings are suggestive for an involvement of the auditory cortex in retrieval attempts of vocalized information and confirm that adopting retrieval orientation is potentially beneficial for retrieval accuracy. The effects of test order on retrieval-related processes might reflect a stronger focus on the newness of items in the more difficult test condition when participants started with this condition.

Quantitative interferometric microscopy with improved full-field phase aberration compensation

Single-shot quantitative interferometric microscopy (QIM) needs a high-accuracy and rapid phase retrieval algorithm. Retrieved phase distributions are often influenced by phase aberration background caused by both imaging system and phase retrieval algorithms. Here, we propose an improved phase aberration compensation (PAC) approach in order to eliminate the phase aberrations inherent in the data. With this method, sample-free parts are identified and used to calculate the background phase, reducing phase errors induced in samples and providing high-quality phase images. We now demonstrate that QIM based on this PAC approach realizes high-quality phase imaging from a single interferogram. This is of great potential for a real-time speedy diagnosis.

Color and Rotated M-Band Dual Tree Complex Wavelet Transform Features for Image Retrieval

In this paper, a novel algorithm which integrates the RGB color histogram and texture features for content based image retrieval. A new set of two- dimensional (2-D) M-band dual tree complex wavelet transform (M_band_DT_CWT) and rotated M_band_DT_CWT are designed to improve the texture retrieval performance. Unlike the standard dual tree complex wavelet transform (DT_CWT), which gives a logarithmic frequency resolution, the M-band decomposition gives a mixture of a logarithmic and linear frequency resolution. Most texture image retrieval systems are still incapable of providing retrieval result with high retrieval accuracy and less computational complexity. To address this problem, we propose a novel approach for image retrieval using M_band_DT_CWT and rotated M_band_DT_CWT (M_band_DT_RCWT) by computing the energy, standard deviation and their combination on each subband of the decomposed image. To check the retrieval performance, two texture databases are used. Further, it is mentioned that the databases used are Brodatz gray scale database and MIT VisTex Color database. The retrieval efficiency and accuracy using proposed features is found to be superior to other existing methods. Index Terms—Color; M-band wavelets; Feature Extraction; M-band dual tree complex wavelets; Image Retrieval.

Similarity metric learning for sketch-based 3D object retrieval

Sketch-Based 3D Object Retrieval (SB3DOR) algorithms retrieve 3D models similar to hand-drawn sketch queries. It is one of the most effective modalities to query 3D models by their shape. However, comparison of a sketch, which is a 2D image, with a 3D model is not straightforward. Most of the SB3DOR algorithms compare a sketch image with images of 3D models rendered from multiple viewpoints. However, retrieval accuracies of state-of-the-art SB3DOR algorithms are still not satisfactory, due, in part, to stylistic variation, semantic influence, abstraction, and drawing error found in sketches. To improve retrieval accuracy of SB3DOR systems, we propose a manifold-based similarity metric learning algorithm that relates two kinds of features, that are, of sketches and 3D models. Features in a high dimensional space often lie on a lower dimensional subspace, or manifold. Feature similarity may be computed more accurately on the manifold than in the original high dimensional space. Our Cross-Domain Manifold Ranking (CDMR) algorithm tries to keep the two distinct feature manifolds, one for sketch features and the other for 3D model features, intact. These two manifolds are interlinked by using inter-feature similarity to form a Cross-Domain Manifold (CDM). If available, semantic labels may also be used in forming the CDM. Relevance diffusion is used to compute similarities between a sketch and 3D models in a database. Experimental evaluation showed that the CDMR algorithm produces higher retrieval accuracy than the algorithms we compared against.

An Effective and Fast Hybrid Framework for Color Image Retrieval

This paper presents a novel, fast and effective hybrid framework for color image retrieval through combination of all the low level features, which gives higher retrieval accuracy than other such systems. The color moment (CMs), angular radial transform descriptor and edge histogram descriptor (EHD) features are exploited to capture color, shape and texture information respectively. A multistage framework is designed to imitate human perception so that in the first stage, images are retrieved based on their CMs and then the shape and texture descriptors are utilized to identify the closest matches in the second stage. The scheme employs division of images into non-overlapping regions for effective computation of CMs and EHD features. To demonstrate the efficacy of this framework, experiments are conducted on Wang’s, VisTex and OT-Scene databases. Inspite of its multistage design, the system is observed to be faster than other hybrid approaches.

Three-dimensional spatiotemporal features for fast content-based retrieval of focal liver lesions.

Content-based image retrieval systems for 3-D medical datasets still largely rely on 2-D image-based features extracted from a few representative slices of the image stack. Most 2 -D features that are currently used in the literature not only model a 3-D tumor incompletely but are also highly expensive in terms of computation time, especially for high-resolution datasets. Radiologist-specified semantic labels are sometimes used along with image-based 2-D features to improve the retrieval performance. Since radiological labels show large interuser variability, are often unstructured, and require user interaction, their use as lesion characterizing features is highly subjective, tedious, and slow. In this paper, we propose a 3-D image-based spatiotemporal feature extraction framework for fast content-based retrieval of focal liver lesions. All the features are computer generated and are extracted from four-phase abdominal CT images. Retrieval performance and query processing times for the proposed framework is evaluated on a database of 44 hepatic lesions comprising of five pathological types. Bull's eye percentage score above 85% is achieved for three out of the five lesion pathologies and for 98% of query lesions, at least one same type of lesion is ranked among the top two retrieved results. Experiments show that the proposed system's query processing is more than 20 times faster than other already published systems that use 2-D features. With fast computation time and high retrieval accuracy, the proposed system has the potential to be used as an assistant to radiologists for routine hepatic tumor diagnosis.

Fusion framework for effective color image retrieval

This paper presents a novel framework for color image retrieval through combination of all the low level features, which gives higher retrieval accuracy. The Color Difference Histogram (CDH) and Angular Radial Transform (ART) features are exploited to capture color, texture and shape information of an image. The CDH algorithm is modified in order to make the proposed system more effective. The proposed fusion framework combines the ranking results of the aforementioned descriptors through various post-classification methods i.e. Borda Count method, Min–max and Z-score normalization. The maximum retrieval accuracy attained in terms of average precision using Min–max normalization on Wang’s database is 78.3% when ART is applied on non-overlapping regions of the images. The proposed fusion framework is recommended because it improves the average retrieval accuracy by approximately 16% and 14% over CDH and ART respectively. Extensive experiments are carried out on different databases to establish the efficacy of the proposed scheme.

Retrieval of urban land surface component temperature using multi-source remote-sensing data

The components of urban surface cover are diversified, and component temperature has greater physical significance and application values in the studies on urban thermal environment. Although the multi-angle retrieval algorithm of component temperature has been matured gradually, its application in the studies on urban thermal environment is restricted due to the difficulty in acquiring urban-scale multi-angle thermal infrared data. Therefore, based on the existing multi-source multi-band remote sensing data, access to appropriate urban-scale component temperature is an urgent issue to be solved in current studies on urban thermal infrared remote sensing. Then, a retrieval algorithm of urban component temperature by multi-source multi-band remote sensing data on the basis of MODIS and Landsat TM images was proposed with expectations achieved in this work, which was finally validated by the experiment on urban images of Changsha, China. The results show that: 1) Mean temperatures of impervious surface components and vegetation components are the maximum and minimum, respectively, which are in accordance with the distribution laws of actual surface temperature; 2) High-accuracy retrieval results are obtained in vegetation component temperature. Moreover, through a contrast between retrieval results and measured data, it is found that the retrieval temperature of impervious surface component has the maximum deviation from measured temperature and its deviation is greater than 1 °C, while the deviation in vegetation component temperature is relatively low at 0.5 °C.