基于极半径曲面矩和HMM的三维模型分类与检索算法

在计算机辅助设计（computer aided design，CAD）中，对机械零件的三维模型进行分类和检索有利于设计人员重用设计信息，从而缩短产品的开发周期，以快速响应市场需求。针对三维模型的分类与检索，提出了一种基于极半径曲面矩和隐马尔科夫模型（hidden Markov model，HMM）的分类与检索算法。首先，计算三维模型的极半径曲面矩并组成特征向量，经排序编码后，将其作为HMM的观测序列；然后，取一部分人工标注过的三维模型作为训练样本，采用添加比例因子的多观测序列B-W（Baum-Welch）算法对HMM进行训练；最后，利用训练好的HMM对三维模型进行分类与检索。实验结果显示，与现有的2种分类与检索算法相比，所提出的算法具有更高的识别率和检索效率。该算法的特点是：极半径曲面矩计算快，不用将三维模型体素化；HMM训练快，分类能力强，且不需要大量训练样本就有一定的分类能力。研究表明，所提出的算法能较好地解决三维模型的分类与检索问题，具有一定的实用价值。     

一种新的基于颜色矢量角的图像空间描述符

在分析利用颜色直方图进行图像检索时存在问题的基础上,提出了一种新的图像空间描述符.首先利用颜色矢量角的特性将图像分为边缘区和平滑区两部分,然后,针对两种区域的不同分布特点,综合利用图像的边缘信息、形状信息及空间分布特性,采用两种不同的描述符--方向颜色矢量角直方图和空间联合分布熵分别来提取图像的边缘区和平滑区的分布特征.实验结果表明,将其用于图像检索,取得了很好的检索效果.     

Feature extraction through generalization of histogram refinement technique for local region‐based object attributes

Content based image retrieval (CBIR) is used to retrieve digital images from large databases. However, the problem of retrieving images on the basis of the contents remains largely unsolved. The proposed method of image retrieval is based on the information provided by histogram analysis of the intensity or grayscale values of images. Some additional properties are also calculated and used that are based on regional characteristics of various objects in the image. The need to retrieve the additional regional properties arises due to the fact that the standard histograms are insensitive to small changes in images. Many images of different types can have similar histograms, because, histograms provide only a coarse characterization of an image. This is the main disadvantage of using histograms. This research is based on the concept of Histogram Refinement (Pass and Zabih, IEEE Workshop Appl Comput Vision ( 1996 ), 96–102). Distributing the grayscale image intensities by splitting the pixels using their intensity values into several classes just like the histogram refinement method can provide an estimate of the object characteristics present in an image. After the calculation of clusters using a color refinement method, the inherent features of each of the clusters is calculated based on the regional properties of the clusters. These additional region based features expound some structural information of the image. Finally, all of these features are used for image retrieval. © 2011 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 21, 298–306, 2011;     

Module-based similarity measurement for commercial aircraft tooling design

Similarity measurement (SM) between pairs of features plays an essential role in the tooling retrieval. However, the number of tooling is usually large and difficult to retrieve. What’s more, the most similar tooling model is not always optimal in the case-based reasoning (CBR). To refine this situation, we propose a novel module-based similarity measurement method. Through dividing each tooling model into different modules, a new similarity is obtained by integrating all the most similar modules (MSMs) from existing tooling models. As a result, the new similarity is superior or at least equal to the result which is derived from previous methods. With the proposed method in this paper, feature similarities of modules between a new tooling ontology and each similar tooling ontology are calculated by fuzzy similarity measurement, and the MSMs are restructured to form a new tooling model after conflict resolution. A case study on forming mould is provided to illustrate the applications and prove the higher similarity by comparing with CBR. The proposed method has been adopted in an aircraft manufacturing company.     

Vesicle sizing by static light scattering: a Fourier cosine transform approach

A Fourier cosine transform method, based on the Rayleigh-Gans-Debye thin-shell approximation, was developed to retrieve vesicle size distribution directly from the angular dependence of scattered light intensity. Its feasibility for real vesicles was partially tested on scattering data generated by the exact Mie solutions for isotropic vesicles. The noise tolerance of the method in recovering unimodal and biomodal distributions was studied with the simulated data. Applicability of this approach to vesicles with weak anisotropy was examined using Mie theory for anisotropic hollow spheres. Aprimitive theory about the first four moments of the radius distribution about the origin, excluding the mean radius, was obtained as an alternative to the direct retrieval of size distributions.     

CANFIS based glioma brain tumor classification and retrieval system for tumor diagnosis

Brain tumor classification and retrieval system plays an important role in medical field. In this paper, an efficient Glioma Brain Tumor detection and its retrieval system is proposed. The proposed methodology consists of two modules as classification and retrieval. The classification modules are designed using preprocessing, feature extraction and tumor detection techniques using Co‐Active Adaptive Neuro Fuzzy Inference System (CANFIS) classifier. The image enhancement can be achieved using Heuristic histogram equalization technique as preprocessing and further texture features as Local Ternary Pattern (LTP) features and Grey Level Co‐occurrence Matrix (GLCM) features are extracted from the enhanced image. These features are used to classify the brain image into normal and abnormal using CANFIS classifier. The tumor region in abnormal brain image is segmented using normalized graph cut segmentation algorithm. The retrieval module is used to retrieve the similar segmented tumor regions from the dataset for diagnosing the tumor region using Euclidean algorithm. The proposed Glioma Brain tumor classification methodology achieves 97.28% sensitivity, 98.16% specificity and 99.14% accuracy. The proposed retrieval system achieves 97.29% precision and 98.16% recall rate with respect to ground truth images.     

基于颜色矢量角度直方图与DCT压缩的鲁棒哈希算法

目的为了改善哈希序列对篡改内容的正确识别率和鲁棒性,提出基于颜色矢量角度直方图耦合离散余弦变换压缩的鲁棒哈希算法。方法结合插值运算和Gaussian滤波器,对图像实施预处理,使其对于任意的认证目标均可产生一个固定长度的哈希序列;引入极坐标变换LPT(Log-polar transform),对规范尺寸图像完成坐标变换,输出二次图像;借助SVD(SingularValueDecompostion)方法来处理二次图像,输出全局鲁棒特征;提取规范尺寸图像中所有像素分量的颜色矢量角度,并计算其对应的直方图,形成色彩感知特征;将全局与色彩感知特征组合,将其视为1D哈希序列;基于离散余弦变换DCT(Discrete Cosine Transform),对1D哈希序列实施压缩,输出一组交流系数;引入Logistic映射,根据其输出的混沌序列来设计差异加密方法,对交流系数实施扩散,从而形成目标哈希序列。基于l2范数距离和优化的识别阈值,对待检测图像内容的真伪实施判别。结果试验结果显示,相对于已有的哈希机制而言,所提方案具备更强的鲁棒性,对各类几何攻击均有更高的检测准确度。结论所提哈希方法具有较高的鲁棒性与识别正确率,对商标检索、信息防伪等领域具有较好的参考价值。     

一种快速图像纹理分析算法

针对目前的纹理分析算法普遍存在计算复杂、运行时间过长的缺点,提出了一种新的图像纹理分析算法。该算法利用差分概念定义出图像的差分矩阵和差分直方图,并从差分直方图中提取特征进行图像纹理的相似度度量。由于本算法在考虑人类视觉特性的基础上尽量采用加减运算,因此不但具有与人类视觉特性相联系的特点,更使得算法的运行效率大大提高。使用本算法对brodatz 图像库进行的检索实验表明,该算法在查准率和检索效率上与共生矩阵算法相比分别提高13.9160%和 9.5 倍,具有与传统算法相比更好的检索效果和更高的检索效率。因此该算法可以满足实时性要求较高的场合,弥补了目前算法运行时间长的缺点。     

Image retrieval using wavelet-based salient regions

Abstract

A content-based image retrieval system normally returns the retrieval results according to the similarity between features extracted from the query image and candidate images. In certain circumstances, however, users may concern more about salient regions in an image of their interest and only wish to retrieve images containing the relevant salient regions while ignoring those irrelevant (such as the background or other regions and objects). Although how to represent the local image properties is still one of the most active research issues, much previous work on image retrieval does not examine salient regions in an image. In this paper, we propose an improved salient point detector based on wavelet transform; it can extract salient points in an image more accurately. Then salient points are segmented into different salient regions according to their spatial distribution. Colour moments and Gabor features of these different salient regions are computed and form a feature vector to index the image. We test the proposed scheme using a wide range of image samples from the Corel Image Library. The experimental results indicate that the method has produced promising results.     

A Cryptograph Domain Image Retrieval Method Based on Paillier Homomorphic Block Encryption

With the rapid development of information network, the computing resources and storage capacity of ordinary users cannot meet their needs of data processing. The emergence of cloud computing solves this problem but brings data security problems. How to manage and retrieve ciphertext data effectively becomes a challenging problem. To these problems, a new image retrieval method in ciphertext domain by block image encrypting based on Paillier homomophic cryptosystem is proposed in this paper. This can be described as follows: According to the Paillier encryption technology, the image owner encrypts the original image in blocks, obtains the image in ciphertext domain, then passes it to the third party server. The server calculates the difference histogram of the image in ciphertext domain according to the public key and establishes the index database. The user passes the retrieved image to the server. The server computes the differential histogram of the retrieved image by public key. Then, compares the similarity of it with the histogram in index database and selects larger similarity images in ciphertext and send them to the user. The user obtains the target image with the private key. The experimental results show that the method is feasible and simple.     

基于颜色信息与空间特征的自适应商标检索算法

目的为了增强商标检索技术对商标特征的描述能力,改善其在外来干扰下的检索精度与鲁棒性。方法提出一种基于颜色与空间特征自适应结合的商标检索算法。首先,引入主颜色描述符(DCD),将其作为颜色特征检测器,并在颜色特征提取时嵌入k-均值聚类算子,增强颜色区域,准确提取颜色特征。随后,每个商标被量化为8个显色的最大值,以便提取每个颜色分量中的空间分布信息。然后,通过利用不同的权重来平衡颜色与空间特征的重要性,定义一种基于模糊直方图分析技术,计算每个商标自适应系数,以准确描述彩色商标的图像特征。最后,通过Euclidean距离进行相似度量,输出检索到的商标。结果实验结果表明,与当前商标检索方法相比,所提算法具有更高的检索精度与鲁棒性,呈现出更理想的P-R曲线,在召回率为0.7时,其检索准确率仍可达到90%。结论文中检索方法具有较高的检索精度,在包装商标检测、商标版权保护等领域中具有良好的应用价值。     

A colour image retrieval scheme based on Z-scanning technique

Abstract

This paper proposes a new colour image retrieval scheme using Z-scanning technique for content-based image retrieval (CBIR). In recent years, the CBIR is a popular research topic for image retrieval. This paper proposes a scheme which employs the Z-scanning technique to extract directional intensity features for measuring the similarity between query and database images. In the multiple channel images, each colour channel can be processed individually or combined into a grey channel Y. In order to extract the features by Z-scanning technique from all images, each channel of all images must be divided into several N×N blocks. In each block, F pairs of pixels are scanned by a ‘Z’ direction to obtain the texture features. Each colour channel can be obtained an M×M Z-scanning co-occurrence matrix (ZSCM) for storing the probability of each relationship of all closest blocks. At the similarity measure stage, the ZSCMs of query image and database images are compared to measure their similarity. The experimental results show that the proposed scheme is beneficial for image retrieval when the images include the same texture or object. On the other hand, the proposed scheme also can get better retrieval results and more efficiency than colour correlogram (CC) technique for colour texture images. Another technique uses motif co-occurrence matrix (MCM) as the feature in similarity measurement. The experimental results show the proposed ZSCM can get better retrieval results and higher recall and precision values than the CC and MCM techniques for public image databases.     

基于字典统计耦合归一化多重距离的图像检索算法

目的针对当前彩色图像检索技术容易受到色彩的干扰,使其鲁棒性不强等不足,提出一种字典统计耦合归一化多重距离的彩色图像检索算法。方法首先将图像量化并转换成一维信号,然后引入字典统计,将一维信号进行字典编码,并计算编码后的图像多样值,在归一化字典距离的基础上嵌入字典编码图像的多样值,从而定义归一化多重距离(NMD)的相似度量准则,利用NMD对查询图像与数据库图像的多样值进行比较与识别,搜索出与查询图像具有相同特征的最相似图像,完成目标检索。结果在COREL数据库的实验结果表明,相对于当前常用的检索技术,该检索算法具有更高的查准率和查全率,可对彩色图像完成精确检索,有效减低了色彩对检索性能的干扰。结论文中算法具有较好的检索精度,能够较好地用于医疗、商标等领域的目标检索。     

基于多层次视觉语义特征融合的图像检索算法

目的为了解决低层特征与中层语义属性间出现的语义鸿沟,以及在将低层特征转化为语义属性的过程中易丢失信息,从而会降低检索精度等问题,设计一种多层次视觉语义特征融合的图像检索算法。方法首先分别提取图像的3种中层特征(深度卷积神经网络(DCNN)特征、Fisher向量、稀疏编码空间金字塔匹配特征(SCSPM));其次,为了对3种特征进行有效融合,定义一种基于图的半监督学习模型,将提取的3个中层特征进行融合,形成一个多层次视觉语义特征,有效结合3种不同中层特征的互补信息,提高图像特征描述,从而降低检索算法中的语义鸿沟;最后,引入具有视觉特性与语义统一的距离函数,根据提取的多层次视觉语义特征来计算查询图像和训练图像的相似度量,完成图像检索任务。结果实验结果表明,与当前检索方法对比,文中算法具有更高的检索精度与效率。结论所提算法具有良好的检索准确度,在医疗、包装商标等领域具有一定的参考价值。     

Shape similarity assessment of polyhedral parts based on boundary models

Shape similarity evaluation is the core technology for many intelligent CAD/CAM applications. In this work a method to evaluate the shape similarity of polyhedral objects based on their boundary models is presented. The proposed method takes the unique boundary models of a query part and a reference part as its input and returns the match coefficient between these two parts. The match coefficient is calculated based on the 'feasible match' of the boundary faces, which is a set of matched faces that preserve the face adjacency relationships. While directly searching for the optimal feasible match in two boundary models could be computationally intractable, a divide-and-conquer method is proposed in this work to find a near-optimal match in a reasonable time. The divide stage corresponds to the extraction and matching of 'sweep features'. A sweep feature of a polyhedral object is a set of parallel faces that are sorted along their common normal vector. The sorted faces in sweep features are matched through an efficient constraint-directed tree search to generate intermediate matches. The intermediate matches that are orientation consistent are merged to form the final feasible matches. A prototype system based on the proposed methodology has been implemented and shape similarity comparison results between various mechanical parts with complicated intersecting features are presented.     

Reversible Data Hiding Based on Run-Level Coding in H.264/AVC Video Streams

This paper presents a reversible data hiding (RDH) method, which is designed by combining histogram modification (HM) with run-level coding in H.264/advanced video coding (AVC). In this scheme, the run-level is changed for embedding data into H.264/AVC video sequences. In order to guarantee the reversibility of the proposed scheme, the last nonzero quantized discrete cosine transform (DCT) coefficients in embeddable 4×4 blocks are shifted by the technology of histogram modification. The proposed scheme is realized after quantization and before entropy coding of H.264/AVC compression standard. Therefore, the embedded information can be correctly extracted at the decoding side. Peak-signal-noise-to-ratio (PSNR) and Structure similarity index (SSIM), embedding payload and bit-rate variation are exploited to measure the performance of the proposed scheme. Experimental results have shown that the proposed scheme leads to less SSIM variation and bit-rate increase.     

Use of sky brightness measurements from ground for remote sensing of particulate polydispersions

The software code SKYEAD.pack for retrieval of aerosol size distribution and optical thickness from data of direct and diffuse solar radiation is described; measurements are carried out with sky radiometers in the wavelength range 0.369-1.048 μm. The treatment of the radiative transfer problem concerning the optical quantities is mainly based on the IMS (improved multiple and single scattering) method, which uses the delta-M approximation for the truncation of the aerosol phase function and corrects the solution for the first- and second-order scattering. Both linear and nonlinear inversion methods can be used for retrieving the size distribution. Improved calibration methods for both direct and diffuse radiation, the data-analysis procedure, the results from the proposed code, and several connected problems are discussed. The results can be summarized as follows: (a) the SKYRAD.pack code can retrieve the columnar aerosol features with accuracy and efficiency in several environmental situations, provided the input parameters are correctly given; (b) when data of both direct and diffuse solar radiation are used, the detectable radius interval for aerosol particles is approximately from 0.03 to 10 μm; (c) besides the retrieval of the aerosol features, the data-analysis procedure also permits the determination of average values for three input parameters (real and imaginary aerosol refractive index, ground albedo) from the optical data; (d) absolute calibrations for the sky radiometer are not needed, and calibrations for direct and diffuse radiation can be carried out with field data; (e) the nonlinear inversion gives satisfactory results in a larger radius interval, without the unrealistic humps that occur with the linear inversion, but the results strongly depend on the first-guess spectrum; (f) aerosol features retrieved from simulated data showed a better agreement with the given data for the linear inversion than for the nonlinear inversion.     

标准硅球直径测量中的光路对准

针对标准单晶硅球直径精密测量的需要,本文在介绍标准硅球直径测量系统原理并分析其光路特点的基础上,根据建立的数学模型,对激光束斜入射标准板时产生的椭圆干涉图像进行了分析,并对不同入射角度时干涉环中心点带来的直径测量误差进行了研究.分析结果显示,在给定的实验条件下,当入射角为10~(-3) rad时,误差已达6.6nm.提出了一种精确调整光束垂直入射平板的方法,实验结果表明,此方法能够使光束入射角的调整优于10~(-5)rad,满足系统测量的要求.     

基于ScSPM-Reranking的高分辨率遥感影像的检索

为了从高分辨率遥感影像中获取详细的地表地物信息,为城市规划、环境监测以及灾情分析提供可靠的数据,进行了高分辨率遥感影像的检索研究,包括对图像的特征提取和图像之间相似度的描述。为了提高图像检索精度,运用了采用稀疏编码(Sc)的空间塔式匹配(Sc SPM)技术和重排序(Reranking)技术,提出了基于Sc SPM结合Reranking(ScSPM-Reranking)的遥感高分辨率影像的检索方法。该方法首先使用Sc SPM提取空间场景的特征,然后结合这些特征使用cityblock距离进行初步检索,最后对初步检索的结果进行Reranking排序,获得高精度的检索结果。同其他检索方法进行了对比实验,实验结果证明,该方法具有较高的检索精度。