Distributed intelligence for multi-camera visual surveillance

Latest advances in hardware technology and state of the art of computer vision and artificial intelligence research can be employed to develop autonomous and distributed monitoring systems. The paper proposes a multi-agent architecture for the understanding of scene dynamics merging the information streamed by multiple cameras. A typical application would be the monitoring of a secure site, or any visual surveillance application deploying a network of cameras. Modular software (the agents) within such architecture controls the different components of the system and incrementally builds a model of the scene by merging the information gathered over extended periods of time. The role of distributed artificial intelligence composed of separate and autonomous modules is justified by the need for scalable designs capable of co-operating to infer an optimal interpretation of the scene. Decentralizing intelligence means creating more robust and reliable sources of interpretation, but also allows easy maintenance and updating of the system. Results are presented to support the choice of a distributed architecture, and to prove that scene interpretation can be incrementally and efficiently built by modular software.     

Two realizations of a general feature extraction framework

A general feature extraction framework is proposed as an extension of conventional linear discriminant analysis. Two nonlinear feature extraction algorithms based on this framework are investigated. The first is a kernel function feature extraction (KFFE) algorithm. A disturbance term is introduced to regularize the algorithm. Moreover, it is revealed that some existing nonlinear feature extraction algorithms are the special cases of this KFFE algorithm. The second feature extraction algorithm, mean-STD¹-norm feature extraction algorithm, is also derived from the framework. Experiments based on both synthetic and real data are presented to demonstrate the performance of both feature extraction algorithms.     

Text information extraction in images and video: a survey

Text data present in images and video contain useful information for automatic annotation, indexing, and structuring of images. Extraction of this information involves detection, localization, tracking, extraction, enhancement, and recognition of the text from a given image. However, variations of text due to differences in size, style, orientation, and alignment, as well as low image contrast and complex background make the problem of automatic text extraction extremely challenging. While comprehensive surveys of related problems such as face detection, document analysis, and image & video indexing can be found, the problem of text information extraction is not well surveyed. A large number of techniques have been proposed to address this problem, and the purpose of this paper is to classify and review these algorithms, discuss benchmark data and performance evaluation, and to point out promising directions for future research.     

Multispectral image compression using eigenregion-based segmentation

In the study, a novel segmentation technique is proposed for multispectral satellite image compression. A segmentation decision rule composed of the principal eigenvectors of the image correlation matrix is derived to determine the similarity of image characteristics of two image blocks. Based on the decision rule, we develop an eigenregion-based segmentation technique. The proposed segmentation technique can divide the original image into some proper eigenregions according to their local terrain characteristics. To achieve better compression efficiency, each eigenregion image is then compressed by an efficient compression algorithm eigenregion-based eigensubspace transform (ER-EST). The ER-EST contains 1D eigensubspace transform (EST) and 2D-DCT to decorrelate the data in spectral and spatial domains. Before performing EST, the dimension of transformation matrix of EST is estimated by an information criterion. In this way, the eigenregion image may be approximated by a lower-dimensional components in the eigensubspace. Simulation tests performed on SPOT and Landsat TM images have demonstrated that the proposed compression scheme is suitable for multispectral satellite image.     

A probabilistic spectral framework for grouping and segmentation

This paper presents an iterative spectral framework for pairwise clustering and perceptual grouping. Our model is expressed in terms of two sets of parameters. Firstly, there are cluster memberships which represent the affinity of objects to clusters. Secondly, there is a matrix of link weights for pairs of tokens. We adopt a model in which these two sets of variables are governed by a Bernoulli model. We show how the likelihood function resulting from this model may be maximised with respect to both the elements of link-weight matrix and the cluster membership variables. We establish the link between the maximisation of the log-likelihood function and the eigenvectors of the link-weight matrix. This leads us to an algorithm in which we iteratively update the link-weight matrix by repeatedly refining its modal structure. Each iteration of the algorithm is a three-step process. First, we compute a link-weight matrix for each cluster by taking the outer-product of the vectors of current cluster-membership indicators for that cluster. Second, we extract the leading eigenvector from each modal link-weight matrix. Third, we compute a revised link weight matrix by taking the sum of the outer products of the leading eigenvectors of the modal link-weight matrices.     

A novel two-phase Hilbert-scan-based search algorithm for block motion estimation using CTF data structure

Motion estimation is one of the kernel issues in MPEG series. In this correspondence, a novel two-phase Hilbert-scan-based search algorithm for block motion estimation is presented. First in the intra-phase, a segmentation of the Hilbert curve is applied to the current block, then a novel coarse-to-fine data structure is developed to eliminate the impossible reference blocks in the search window of the reference frame. In the inter-phase, a new prediction scheme for estimating the initial motion vector of the current block is presented. Experimental results reveal that when compared to the GAPD algorithm, our proposed algorithm has better execution time and estimation accuracy performance. Under the same estimation accuracy, our proposed algorithm has better execution time performance when compared to the FS algorithm. In addition, when comparing with the TSS algorithm, our proposed algorithm has better estimation accuracy performance, but has worse execution time performance.     

Color balancing of digital photos using simple image statistics

The great diffusion of digital cameras and the widespread use of the internet have produced a mass of digital images depicting a huge variety of subjects, generally acquired by unknown imaging systems under unknown lighting conditions. This makes color balancing, recovery of the color characteristics of the original scene, increasingly difficult. In this paper, we describe a method for detecting and removing a color cast (i.e. a superimposed color due to lighting conditions, or to the characteristics of the capturing device), from a digital photo without any a priori knowledge of its semantic content. First a cast detector, using simple image statistics, classifies the input images as presenting no cast, evident cast, ambiguous cast, a predominant color that must be preserved (such as in underwater images or single color close-ups) or as unclassifiable. A cast remover, a modified version of the white balance algorithm, is then applied in cases of evident or ambiguous cast. The method we propose has been tested with positive results on a data set of some 750 photos.     

People tracking based on motion model and motion constraints with automatic initialization

Human motion analysis is currently one of the most active research topics in computer vision. This paper presents a model-based approach to recovering motion parameters of walking people from monocular image sequences in a CONDENSATION framework. From the semi-automatically acquired training data, we learn a motion model represented as Gaussian distributions, and explore motion constraints by considering the dependency of motion parameters and represent them as conditional distributions. Then both of them are integrated into a dynamic model to concentrate factored sampling in the areas of the state-space with most posterior information. To measure the observation density with accuracy and robustness, a pose evaluation function (PEF) combining both boundary and region information is proposed. The function is modeled with a radial term to improve the efficiency of the factored sampling. We also address the issue of automatic acquisition of initial model pose and recovery from severe failures. A large number of experiments carried out in both indoor and outdoor scenes demonstrate that the proposed approach works well     

Connectivity oriented fast Hough transform for tool wear monitoring

Tool wear monitoring can be achieved by analyzing the texture of machined surfaces. In this paper, we present the new connectivity oriented fast Hough transform, which easily detects all line segments in binary edge images of textures of machined surfaces. The features extracted from line segments are found to be highly correlated to the level of tool wear. A multilayer perceptron neural network is applied to estimate the flank wear in various machining processes. Our experiments show that this Hough transform based approach is effective in analyzing the quality of machined surfaces and could be used to monitor tool wear. A performance analysis of our Hough transform is also provided.