The choice of vantage objects for image retrieval

Suppose that we have a matrix of dissimilarities between n images of a database. For a new image, we would like to select the most similar image of our database. Because it may be too expensive to compute the dissimilarities for the new object to all images of our database, we want to find p?n “vantage objects” (Pattern Recognition 35 (2002) 69) from our database in order to select a matching image according to the least Euclidean distance between the vector of dissimilarities between the new image and the vantage objects and the corresponding vector for the images of the database. In this paper, we treat the choice of suitable vantage objects. We suggest a loss measure to assess the quality of a set of vantage objects: For every image, we select a matching image from the remaining images of the database by use of the vantage set, and we average the resulting dissimilarities. We compare two classes of choice strategies: The first one is based on a stepwise forward selection of vantage objects to optimize the loss measure. The second is to choose objects as representative as possible for the whole range of the database.     

Recognition of multiple configurations of objects with limited data

One of the main goals of image understanding and computer vision applications is to recognize an object from various images. A lot of studies on recognizing objects based on invariable shapes have been explored, however, in reality, there are many objects with multiple configurations, which are very difficult to be recognized. We call this kind of problem as the recognition of multiple configurations of objects (RMCO). To achieve RMCO, firstly we obtain a shortest path (the Geodesic distance path) between two feature vectors in pre-shape spaces; along this obtained path, we can generate a series of data which can be used to recognize the observed objects by using shape space theories. In other words, we may augment the database content with very limited data to recognize more objects.     

Efficient image retrieval through vantage objects

We describe a new indexing structure for general image retrieval that relies solely on a distance function giving the similarity between two images. For each image object in the database, its distance to a set of m predetermined vantage objects is calculated; the m-vector of these distances specifies a point in the m-dimensional vantage space. The database objects that are similar (in terms of the distance function) to a given query object can be determined by means of an efficient nearest-neighbor search on these points. We demonstrate the viability of our approach through experimental results obtained with two image databases, one consisting of about 5200 raster images of stamps, the other containing about 72,000 hieroglyphic polylines.     

Generalized distances in digital geometry

A generalized distance measure called m-neighbor distance in n-D quantized space is presented. Its properties as a metric are examined. It is shown to give the shortest path length between two points in n-D digital space. An algorithm for finding such a shortest path between two points is presented. It is shown that lower dimension (2-D and 3-D) distance measures presently used in digital geometry can easily be derived as special cases. Other properties of m-neighbor distance are also examined.     

An active learning framework for content-based information retrieval

We propose a general active learning framework for content-based information retrieval. We use this framework to guide hidden annotations in order to improve the retrieval performance. For each object in the database, we maintain a list of probabilities, each indicating the probability of this object having one of the attributes. During training, the learning algorithm samples objects in the database and presents them to the annotator to assign attributes. For each sampled object, each probability is set to be one or zero depending on whether or not the corresponding attribute is assigned by the annotator. For objects that have not been annotated, the learning algorithm estimates their probabilities with biased kernel regression. Knowledge gain is then defined to determine, among the objects that have not been annotated, which one the system is the most uncertain. The system then presents it as the next sample to the annotator to which it is assigned attributes. During retrieval, the list of probabilities works as a feature vector for us to calculate the semantic distance between two objects, or between the user query and an object in the database. The overall distance between two objects is determined by a weighted sum of the semantic distance and the low-level feature distance. The algorithm is tested on both synthetic databases and real databases of 3D models. In both cases, the retrieval performance of the system improves rapidly with the number of annotated samples. Furthermore, we show that active learning outperforms learning based on random sampling.     

Towards modeling the query processing relevant shape complexity of 2D polygonal spatial objects

The shape complexity of two-dimensional (2D) polygonal spatial objects has implications for how the object can be best represented in a spatial database, and for the query-processing performance of that object. Nevertheless few useful definitions of query-processing relevant spatial complexity are available. A query-processing oriented shape complexity measure is likely to be different from a fractal measure of shape complexity that focused on compression/decompression or a shape complexity measure that would be used for pattern recognition, and should give better performance for the analysis of query processing. It could be used to classify spatial objects, cluster spatial objects in multiprocessor database systems. In a recent paper Brinkhoff et al. (T. Brinkhoff, H-P. Kriegel, R. Shneider, A. Braun, Measuring the complexity of spatial objects, Proceedings of the 3rd ACM International Workshop on Advances in Geographic Information Systems, Baltimore, MD, 1995, pp. 109–117) demonstrated the usefulness of a spatial complexity measure. They did not, however, offer much theoretical justification for their choice of parameters nor for the functional form that they used. In this paper we present a conceptual framework for discussing the query processing oriented shape complexity measures for spatial objects. It is hoped that this will lead to the development of improved measures of spatial complexity.     

DART+: Direction-aware bichromatic reverse k nearest neighbor query processing in spatial databases

This article presents a novel type of queries in spatial databases, called the direction-aware bichromatic reverse k nearest neighbor(DBRkNN) queries, which extend the bichromatic reverse nearest neighbor queries. Given two disjoint sets, P and S, of spatial objects, and a query object q in S, the DBRkNN query returns a subset P′ of P such that k nearest neighbors of each object in P′ include q and each object in P′ has a direction toward q within a pre-defined distance. We formally define the DBRkNN query, and then propose an efficient algorithm, called DART, for processing the DBRkNN query. Our method utilizes a grid-based index to cluster the spatial objects, and the B⁺-tree to index the direction angle. We adopt a filter-refinement framework that is widely used in many algorithms for reverse nearest neighbor queries. In the filtering step, DART eliminates all the objects that are away from the query object more than a pre-defined distance, or have an invalid direction angle. In the refinement step, remaining objects are verified whether the query object is actually one of the k nearest neighbors of them. As a major extension of DART, we also present an improved algorithm, called DART+, for DBRkNN queries. From extensive experiments with several datasets, we show that DART outperforms an R-tree-based naive algorithm in both indexing time and query processing time. In addition, our extension algorithm, DART+, also shows significantly better performance than DART.     

Automatic key video object plane selection using the shapeinformation in the MPEG-4 compressed domain

Object-based video representation, such as the one suggested by the MPEG-4 standard, offers a framework that is better suited for object-based video indexing and retrieval. In such a framework, the concept of a “key frame” is replaced by that of a “key video object plane”. In this paper, we propose a method for key video object plane selection using the shape information in the MPEG-4 compressed domain. The shape of the video object (VO) is approximated using the shape coding modes of I, P, and B video object planes (VOPs) without decoding the shape information in the MPEG-4 bit stream. Two popular shape distance measures, the Hamming and Hausdorff distance measures, are modified to measure the similarities between the approximated shapes of the video objects. Although they feature different computational and implementation complexity tradeoffs, the corresponding algorithms achieve essentially the same performance levels in selecting key video object planes that represent efficiently the salient content of the video objects     

Continuous reverse k nearest neighbor monitoring on moving objects in road networks

Continuous reverse k nearest neighbor (CRkNN) monitoring in road networks has recently received increasing attentions. However, there is still a lack of efficient CRkNN algorithms in road networks up to now. In road networks, moving query objects and data objects are restricted by the connectivity of the road network and both the object–query distance and object–object distance updates affect the result of CRkNN queries. In this paper, we present a novel algorithm for continuous and incremental evaluation of CRkNN queries in road networks. Our method is based on a novel data structure called dual layer multiway tree (DLM tree) we proposed to represent the whole monitoring region of a CRkNN query q. We propose several lemmas to reduce the monitoring region of q and the number of candidate objects as much as possible. Moreover, by associating a variable NN_count with each candidate object, we can simplify the monitoring of candidate objects. There are a large number of objects roaming in a road network and many of them are irrelevant to a specific CRkNN query of a query object q. To minimize the processing extension, for a road in the network, we give an IQL list and an IQCL list to specify the set of query objects and data objects whose location updates should be maintained for CRkNN processing of query objects. Our CRkNN method consists of two phase: the initial result generating phase and incremental maintenance phase. In each phase, algorithms with high performance are proposed to make our CRkNN method more efficient. Extensive simulation experiments are conducted and the result shows that our proposed approach is efficient and scalable in processing CRkNN queries in road networks.     

Capture local information in shape representation

Research in content-based image retrieval has been around for over a decade. While the research community has successfully exploited content features such as color and texture, finding an effective shape representation and measure remains a challenging task. The shape feature is particularly crucial for the success of content-based systems as it carries meaningful semantics of the objects of interest and fits more naturally into humans’ perception of similarity. In this paper, we present our approach to use the shape feature for image retrieval. First, we introduce an effective image decomposition method called Crawling Window (CW) to distinguish the outline of each object in the image. Second, to represent each individual shape, we propose a novel representation model called component Distance Distribution Function and its measure. Traditionally, an object is represented by a set of points on the shape’s contour. Our idea is to first compute the distance between each point and the center of the object. The distance values for all points form a signal, which we call Distance Distribution Function (DDF). Each DDF is then divided into component DDFs (cDDF) by taking local signal information into account. Finally, a transformation technique is employed to generate the feature vector for each cDDF. All vectors from the cDDFs in circular order construct the final shape representation. The model is invariant to position, scaling, rotation and starting point. The similarity measure model based on the new representation is also introduced. Our extensive experiments show that our models are more effective than the existing representation model, both in the shape and the image level.

Optimal piecewise linear motion of an object among obstacles

We present an algorithm for determining the shortest restricted path motion of a polygonal object amidst polygonal obstacles. The class of motions which are allowed can be described as follows: a designated vertex,P, of the polygonal object traverses a piecewise linear path, whose breakpoints are restricted to the vertices of the obstacles. The distance measure being minimized is the length of the path traversed byP. Our algorithm runs in timeO(n ⁴kogn). We also discuss a variation of this algorithm which minimizes any positive linear combination of length traversed byP and angular rotation of the ladder aboutP. This variation requiresO(n ₅) time.     

基于距离比上下文的形状描述与识别方法

基于形状轮廓上的采样点到形状质心的距离,提出了一种距离比上下文形状描述符,用于形状识别和检索。该描述符计算简单,能有效区分不同形状,本质上具有平移、缩放不变性,且在一定程度上能杭部分遮挡和形变。用动态规划算法度量形状比上下文之间的距离,解决了对起始轮廓点的选择问题。在kimia' s-99形状图像数据库中的实验结果表明,该方法在单目标封闭轮廓的形状图像检索中取得了良好的效果。     

A multi-resolution surface distance model for <Emphasis Type="Italic">k</Emphasis>-NN query processing

A spatial k-NN query returns k nearest points in a point dataset to a given query point. To measure the distance between two points, most of the literature focuses on the Euclidean distance or the network distance. For many applications, such as wildlife movement, it is necessary to consider the surface distance, which is computed from the shortest path along a terrain surface. In this paper, we investigate the problem of efficient surface k-NN (sk-NN) query processing. This is an important yet highly challenging problem because the underlying environment data can be very large and the computational cost of finding the shortest path on a surface can be very high. To minimize the amount of surface data to be used and the cost of surface distance computation, a multi-resolution surface distance model is proposed in this paper to take advantage of monotonic distance changes when the distances are computed at different resolution levels. Based on this innovative model, sk-NN queries can be processed efficiently by accessing and processing surface data at a just-enough resolution level within a just-enough search region. Our extensive performance evaluations using real world datasets confirm the efficiency of our proposed model.     

ALSBIR: A local-structure-based image retrieval

The general-purpose shape retrieval problem is a challenging task. Particularly, an ideal technique, which can work in clustered environment, meet the requirements of perceptual similarity measure on partial query and overcoming dimensionality curse and adverse environment, is in demand. This paper reports our study on one local structural approach that addresses these issues. Shape representation and indexing are two key points in shape retrieval. The proposed approach combines a novel local-structure-based shape representation and a new histogram indexing structure. The former makes possible partial shape matching of objects without the requirement of segmentation (separation) of objects from complex background, while the latter has an advantage on indexing performance. The search time is linearly proportional to the input complexity. In addition, the method is relatively robust under adverse environments. It is able to infer retrieval results from incomplete information of an input by first extracting consistent and structurally unique local neighborhood information from inputs or models, and then voting on the optimal matches. Thousands of images have been used to test the proposed concepts on sensitivity analysis, similarity-based retrieval, partial query and mixed object query. Very encouraging experimental results with respect to efficiency and effectiveness have been obtained.     

Spatial inverse query processing

Traditional spatial queries return, for a given query object q, all database objects that satisfy a given predicate, such as epsilon range and k-nearest neighbors. This paper defines and studies inverse spatial queries, which, given a subset of database objects Q and a query predicate, return all objects which, if used as query objects with the predicate, contain Q in their result. We first show a straightforward solution for answering inverse spatial queries for any query predicate. Then, we propose a filter-and-refinement framework that can be used to improve efficiency. We show how to apply this framework on a variety of inverse queries, using appropriate space pruning strategies. In particular, we propose solutions for inverse epsilon range queries, inverse k-nearest neighbor queries, and inverse skyline queries. Furthermore, we show how to relax the definition of inverse queries in order to ensure non-empty result sets. Our experiments show that our framework is significantly more efficient than naive approaches.