A novel local octa-pattern feature descriptor for image retrieval

The local tetra patterns (LTrPs) gives four-directional information and ignores the diagonal pixel information, thereby affecting the retrieved image efficiency. In the present work, a novel retrieval approach has been proposed using local octa-patterns (LOcPs) for content-based image indexing and retrieval. The proposed approach encodes the center pixel directional information with its eight adjacent neighbors, from the directions that are computed using the first-order derivatives. Also the nth-order LOcP is computed using \((n-1)\)th-order local direction variations. In addition, the performance of the developed method by combining it with the Gabor transform has been analyzed. The performance of the proposed technique has been compared to existing techniques like LBP, LTP, LDP, and LTrP on Corel-1000 database (DB1) and Describable Textures Dataset (DB2). The performance observed shows that the developed method improves the retrieval parameters from 75.9%/77.13% to 79.4%/81.5% in the form of average precision on DB1/DB2 databases.     

Statistical texture retrieval in noise using complex wavelets

This paper investigates the use of complex wavelets for statistical texture retrieval in a noisy environment, in which the query image is contaminated by noise. To account for the presence of noise, the feature extraction step is based on parameter estimation in noise where features are extracted from the noisy query image by modeling the magnitude and phase of complex subband coefficients of the clean image, and relating the model's parameters to the noisy coefficients. In addition to using only the magnitude or phase which is in the form of the relative phase, we incorporate both magnitude and phase information to further improve the accuracy rate. The simulation results show the retrieval rate improvement by estimating the clean parameters from the noisy query image instead of assuming that the query image is clean. Furthermore, using both magnitude and phase of complex coefficients improves the accuracy rate from using either magnitude or phase alone, and that using complex-valued wavelets yields higher rate than using real-valued wavelets.     

Content-based image retrieval using a fusion of global and local features

Color, texture, and shape act as important information for images in human recognition. For content-based image retrieval, many studies have combined color, texture, and shape features to improve the retrieval performance. However, there have not been many powerful methods for combining all color, texture, and shape features. This study proposes a content-based image retrieval method that uses the combined local and global features of color, texture, and shape. The color features are extracted from the color autocorrelogram; the texture features are extracted from the magnitude of a complete local binary pattern and the Gabor local correlation revealing local image characteristics; and the shape features are extracted from singular value decomposition that reflects global image characteristics. In this work, an experiment is performed to compare the proposed method with those that use our partial features and some existing techniques. The results show an average precision that is 19.60% higher than those of existing methods and 9.09% higher than those of recent ones. In conclusion, our proposed method is superior over other methods in terms of retrieval performance.     

基于Multibandelets的自适应图像压缩

为了在图像压缩时更好地保护具有方向性的几何结构信息,该文构建了一种新的基函数,称为Multibandelets,并结合Shannon编码用于自然图像的压缩。实验结果表明:与多小波、具有同样消失矩的小波和Bandelets相比较,基于Multibandelets的图像压缩在视觉效果和客观衡量指标两方面都有改善,尤其对具有方向性的细节和纹理信息具有更好的表示。     

PicToSeek: combining color and shape invariant features for imageretrieval

We aim at combining color and shape invariants for indexing and retrieving images. To this end, color models are proposed independent of the object geometry, object pose, and illumination. From these color models, color invariant edges are derived from which shape invariant features are computed. Computational methods are described to combine the color and shape invariants into a unified high-dimensional invariant feature set for discriminatory object retrieval. Experiments have been conducted on a database consisting of 500 images taken from multicolored man-made objects in real world scenes. From the theoretical and experimental results it is concluded that object retrieval based on composite color and shape invariant features provides excellent retrieval accuracy. Object retrieval based on color invariants provides very high retrieval accuracy whereas object retrieval based entirely on shape invariants yields poor discriminative power. Furthermore, the image retrieval scheme is highly robust to partial occlusion, object clutter and a change in the object's pose. Finally, the image retrieval scheme is integrated into the PicToSeek system on-line at http://www.wins.uva.nl/research/isis/PicToSeek/ for searching images on the World Wide Web.     

Adaptive directional lifting-based wavelet transform for image coding.

We present a novel 2-D wavelet transform scheme of adaptive directional lifting (ADL) in image coding. Instead of alternately applying horizontal and vertical lifting, as in present practice, ADL performs lifting-based prediction in local windows in the direction of high pixel correlation. Hence, it adapts far better to the image orientation features in local windows. The ADL transform is achieved by existing 1-D wavelets and is seamlessly integrated into the global wavelet transform. The predicting and updating signals of ADL can be derived even at the fractional pixel precision level to achieve high directional resolution, while still maintaining perfect reconstruction. To enhance the ADL performance, a rate-distortion optimized directional segmentation scheme is also proposed to form and code a hierarchical image partition adapting to local features. Experimental results show that the proposed ADL-based image coding technique outperforms JPEG 2000 in both PSNR and visual quality, with the improvement up to 2.0 dB on images with rich orientation features.     

基于显著性区域和蚁群算法的图像检索研究

针对提取到的图像特征受背景信息干扰,不能有针对性地提取到所需要的图像信息影 响检索精度。为了解决这一问题,本文提出一种基于改进VGGNet(visual geometry group network)和蚁群算法的图像显著性区 域检索算法。首先,利用类激活映射(class activation mapping, CMA)算法对图像显著性区域进行提取,剔除图像背景信息;然后使 用训练好的RS-VGG16模型提取图像显著性区域特征来表征图像。引入主成分分析(principal component analysis, PCA)算法,对高维特征 进行降维的同时减少特征信息的损失。最后,引入蚁群算法对检索结果进行优化。在corel_ 5000数据集上,选取基于VGG16网络的图像全局特征检索算法以及传统的BOF (bag of features)图像检索算法进 行对比试验。本文提出算法相较于基于VGG16网络的图像检索算法,平均查准率(mean average precision, MAP)值平均提升约4.36% ,相较于传统的BOF算法,MAP值平均提升约16.99%。实验结果表明本 文提出算法能够很好地去除图像背景信息的干扰,具有更优的检索性能。     

基于多尺度相位特征的图像检索方法

在基于内容的图像检索中,一个关键的问题是图像视觉内容的表述。而传统的颜色,形状和纹理特征对于图像内容的表述尚且不够完备。为进一步提高检索准确率,针对人眼视觉特性,该文提出了一种基于多尺度相位特征的图像检索方法。该方法首先采用尺度空间理论得到图像的多尺度描述,然后通过复数可调滤波(complex steerable filtering)提取图像的多尺度相位信息并利用直方图投影获取全局统计的多尺度相位特征。在通用数据库COREL 5000上的实验结果表明,该特征相对经典的颜色特征提高至少5%检索准确率,且能对之提供有效补充。     

基于NSCT及熵的旋转不变彩色图像检索算法

为了解决图像在转载过程中所产生的旋转变化和尺 度变化对检索的影响,根据熵的对称性,提出了基于NSCT及熵的旋转不变图像检索算法。首 先,利用非下采样轮廓波变换(NSCT)对图像进行多尺度、多方向分解,对不同尺度、同方 向的高频方向子带求多尺度积,以减小尺度变 化和噪声对检索效率的影响;然后,考虑到图像旋转后各方向子带在整幅图像中的能量比例 不会发生变化, 将各方向子带的能量比例作为概率矢量,各方向子带的粗糙度作为权值求取图像的加权信息 熵,作为具 有旋转不变性的图像纹理特征,利用矩提取图像的颜色和形状特征;最后,归一化3种特征 来比较两幅图 像的相似性。性能测试表明,本文所提出的方法对旋转变换鲁棒性强,且具有很高的查准率 和查全率。     

MRI and CT image indexing and retrieval using local mesh peak valley edge patterns

In this paper, a new pattern based feature, local mesh peak valley edge pattern (LMePVEP) is proposed for biomedical image indexing and retrieval. The standard LBP extracts the gray scale relationship between the center pixel and its surrounding neighbors in an image. Whereas the proposed method extracts the gray scale relationship among the neighbors for a given center pixel in an image. The relations among the neighbors are peak/valley edges which are obtained by performing the first-order derivative. The performance of the proposed method (LMePVEP) is tested by conducting two experiments on two benchmark biomedical databases. Further, it is mentioned that the databases used for experiments are OASIS−MRI database which is the magnetic resonance imaging (MRI) database and VIA/I–ELCAP-CT database which includes region of interest computer tomography (CT) images. The results after being investigated show a significant improvement in terms average retrieval precision (ARP) and average retrieval rate (ARR) as compared to LBP and LBP variant features.     

基于兴趣点颜色形状特征的图像检索方法

提出了一种基于兴趣点的图像检索新方法ACSM（Annular Color and Shape Moment）．利用兴趣点的位置信息，提取兴趣点周围局部区域的颜色矩特征和区域形状不变矩特征。ACSM算法不仅对平穆、旋转具有良好的鲁棒性，而且克服了传统颜色矩和形状不变矩没有位置信忠的缺陷。实验结果表明，由于将颜色和形状特征有机结合，ACSM算法有效提高了图像检索的准确性。     

采用卷积神经网络的红外和可见光图像块匹配

红外和可见光图像块匹配在视觉导航和目标识别等任务中有着广泛的应用。由于红外和可见光传感器有不同的成像原理,红外和可见光图像块匹配更加具有挑战。深度学习在可见光领域图像的块匹配上取得了很好的性能,但是它们很少涉及到红外和可见光的图像块。文中提出了一种基于卷积神经网络的红外和可见光的图像块匹配网络。此网络由特征提取和特征匹配两部分组成。在特征提取过程中,使用对比和三重损失函数能够最大化不同类的图像块的特征距离,缩小同一类图像块的特征距离,使得网络能够更加关注于图像块的公共特征,而忽略红外和可见光成像之间差异。在红外和可见光图像中,不同尺度的空间特征能够提供更加丰富的区域和轮廓信息。红外和可见光图像块的高层特征和底层特征融合可以有效地提升特征的表现能力。改进后的网络相比于先前卷积神经匹配网络,准确率提升了9.8%。     

基于融合特征的现勘图像检索结果优化算法

刑侦现勘图像数据库是具有保密性高、图像内容罕见等极具行业特色的图像数据库.针对现勘图像内容复杂、目标物体不明确的特点,提出了DCT-DCT波纹理特征,并与HSV颜色直方图特征、GIST特征相融合构成融合特征.与常用的图像特征相比,DCT-DCT波纹理特征能够得到较高的检索效率,而融合特征的平均检索查准率高于构成其本身的三种特征的平均检索查准率.最后,将语义分析技术引入到检索过程中,提出基于检索结果优化的现勘图像检索算法,利用支持向量机（Support Vector Machine,SVM）分类器对查询图像进行语义提取,并对初次检索的结果进行语义分析,根据初检结果中语义类别的占比选择二次检索方案,该算法能在按例查询的基础上进一步提高平均检索查准率.