Efficient Mean-Shift Tracking via a New Similarity Measure

In order to enhance the robustness of IR fast small target tracking, a novel mean shift tracking algorithm using improved similarity measure of is proposed. Firstly, problems of local background interfering in original mean shift algorithm for tracking fast motion small target is analyzed, and the reasons is located in the Bhattacharyya coefficient similarity measure expression for all gray weights of components are same, which cannot reflect the advantage contribution of the small target’s gray component in the process of measuring similarity, causing serious interference of the background in the tracking process, leaving the algorithm converging easily. Therefore, to solve this problem, the improvements Bhattacharyya coefficient similarity measure with the local background information fused is proposed. Then, shift vector is deduced in the framework of mean shift by regarding Bhattacharyya coefficients as the similarity measure.In shifting process, the robustness of the small target tracking is improved effectively according to target gray level of large membership degree with high shift weight, and vice versa with low shift weight, which the background interference is suppressed to some extent. In sake of verifying the performance of the proposed algorithm, the classical mean shift algorithm and the algorithm of this paper is used in the target tracking simulation experiment, as well as the infrared image sequences containing the small fast targets of uncooled infrared camera is used. Finally the experimental result indicates that the performance of tracking the small fast target in IR images is robust.

Enhancing color image retrieval performance with feature fusion and non-linear support vector machine classifier

In this paper, we improve the real-time object tracking algorithm of Yang [1] which uses a symmetric similarity function between spatially smoothed kernel-density estimates of the model and the target distributions. This spatial smoothed process applied on the centre points of the probability density functions increases not only computational complexity but also noise sensitivity. After reducing background information and using a new simple similarity function between the model and the target distributions, the proposed algorithm successfully coped with camera motion, partial occlusions and clutter with lower computational complexity. For tracking object, the mean shift algorithm is used iteratively. The simplicity of the new similarity function leads to an efficient and robust nonparametric tracking algorithm. The mathematical results obtained about the performance of the proposed algorithm on several image sequences are represented comparatively in a tabular format.

Infrared object tracking plays a key role in many research fields, and there is a series of work on applying particle filter to this tracking problem. Most of the PF-based tracking algorithms utilize the Bhattacharyya coefficient as a similarity measure, however, its performance in infrared object tracking is limited due to insufficient discriminative power. In this paper, we present a combined similarity measure under the particle filter framework, which integrates the advantages of the Bhattacharyya coefficient, histogram intersection, and structural similarity. The experimental results are gained by using different infrared image sequences, which show that the proposed measure gives superior discriminative power and achieves more robust and stable tracking performance than the traditional approach.

In traditional Mean Shift algorithm, color histogram is usually used as the features vectors and the dissimilarity between the referenced targets and the target candidates is expressed by the metric derived from the Bhattacharyya coefficients. The traditional Mean Shift procedure is used to find the real position of the target by looking for the regional minimum of the distance function iteratively. While the target's color is similar to the background, the algorithm will miss the target. This paper presents a new mean shift algorithm based on spatial edge orientation histograms, using space distribution and texture information as matching information. Meanwhile a Kalman filter will be used to predict the target's position. Experimental results demonstrate that the proposed algorithm can deal with intricate conditions, such as significant clutter, partial occlusions, and it can track objects efficiently and robustly.

The mean-shift algorithm has achieved considerable success in object tracking due to its simplicity and efficiency.Color histogram is a common feature in the description of an object.However, the kernel-based color histogram may not have the ability to discriminate the object from clutter background.To boost the discriminating ability of the feature, based on background contrasting, this letter presents an improved Bhattacharyya similarity metric for mean-shift tracking.Experiments show that the proposed tracker is more robust in relation to background clutter.

The mean‐shift method is widely used to locate a target object quickly in sequential images. The mean‐shift algorithm takes advantage of a colour distribution with a uniform quantisation. However, the quantisation method ignores the close relationship of colour statistics. The uniform distribution also results in a colour histogram with many empty bins, which introduces additional computation cost in the tracking procedure. To reduce the number of these redundant, empty bins, the authors present a new optimal colour‐based, mean‐shift algorithm for tracking objects. In the proposed method, the optimal colours are extracted by a histogram agglomeration, which clusters three‐dimensional (3D) colour histogram bins with the frequency ratios of 3D colour values. After obtaining optimal colours in a RGB colour histogram, the target image is represented by the indices of the optimal colours. The mean‐shift algorithm thus creates a confidence map in a candidate image based on the optimal colour histogram in the target image. It then finds the peak of the confidence map near the previous position of an object area. Comparative experiments with the conventional mean‐shift method showed that our method has the advantages of decreased processing time and improved tracking accuracy.

The color histogram is a statistical behavior for robust pattern search or matching; however, difficulties have arisen in using it to discriminate among similar objects. Our method, called absent color indexing (ABC), describes how to use absent or minor colors as a feature in order to solve problems while robustly recognizing images, even those with similar color features. The proposed approach separates a source color histogram into apparent (AP) and absent (AB) color histograms in order to provide a fair way of focusing on the major and minor contributions together. A threshold for this separation is automatically obtained from the mean color histogram by considering the statistical significance of the absent colors. After these have been separated, an inversion operation is performed to reinforce the weight of AB. In order to balance the contributions of the two histograms, four similarity measures are utilized as candidates for combination with ABC. We tested the performance of ABC in terms of the F-measure using different similarity measures, and the results show that it is able to achieve values greater than 0.95. Experiments on Mondrian random patterns verify the ability of ABC to distinguish similar objects by margin. The results of extensive experiments on real-world images and open databases are presented here in order to demonstrate that the performance of our relatively simple algorithm remained robust even in difficult cases.

This paper addresses the problem of small scale target tracking. The divided-by-zero problem in the weight computation of mean shift algorithm and its associated tracking interrupt problem are presented. To tackle these problems, the Parzen window density estimation method is modified to interpolate the histogram of the target candidate. Then the Kullback-Leibler distance is employed as a new similarity measure between the target model and the target candidate. Its corresponding weight computation and new location expressions are derived. On the basis of these works, we propose a new small target tracking algorithm using mean shift framework. The tracking experiments for real world video sequences show that the proposed algorithm can track the target successively and accurately. It can successfully track very small targets with only 6×12 pixels.

In order to solve the problems of motion blur and fast motion, a new robust object tracking algorithm using the Kernelized Correlation Filters (KCF) and the Mean Shift (MS) algorithm, called KCFMS is presented in this paper. The object tracking process can be described as: First, we give the initial position and size of the object and use the Mean Shift algorithm to obtain the position of the object. Second, the Kernelized Correlation Filtering algorithm is used to obtain the position of the object in the same frame. Third, we use the cross update strategy to update the object models. In order to improve the tracking speed as much as possible, our object tracking algorithm works only over one layer. This hybrid algorithm has a good tracking effect on the target fast motion and motion blur. We present extensive experimental results on a number of challenging sequences in terms of efficiency, accuracy and robustness.

A superpixels based interactive image segmentation algorithm is proposed in this paper. Firstly the initial segmentation is obtained by MeanShift algorithm, and then a graph is built using pre-segmented regions as nodes, finally min-cut/maxflow algorithm is implemented for global solution. In this process, each region is represented by a color histogram and Bhattacharyya coefficient is chosen to calculate the similarity between any two regions. Extensive experiments are performed and the results show that the presented algorithm obtains much more satisfactory segmentation results with less user interaction and less comsuming time than MSRM algorithm.

In a content-based image retrieval (CBIR) system, indexing feature vectors and the similarity measure between feature vectors are two key factors for retrieval performance. We present a new CBIR system with statistical-model based image feature extraction in the wavelet domain and a Kullback divergence based similarity measure. A two component Gaussian mixture model (GMM) in the wavelet domain is employed and the model parameters are used to form features for image indexing. A new Kullback divergence based similarity measure is then presented for image retrieval. The experimental results demonstrate that the similarity measure based on the Kullback divergence is more effective than conventional similarity measures, such as the city-block distance and the Euclidean distance. It is shown that the new CBIR system, with the combination of the GMM and the new Kullback divergence based similarity measure, outperforms most other methods in retrieval performance for texture images, while keeping a comparable level of computational complexity.

In this paper, we define a similarity measure between images in the context of (indexing and) retrieval. We use the Kullback-Leibler (KL) divergence to compare sparse multiscale image representations. The KL divergence between parameterized marginal distributions of wavelet coefficients has already been used as a similarity measure between images. Here we use the Laplacian pyramid and consider the dependencies between coefficients by means of nonparametric distributions of mixed intra/interscale and interchannel patches. To cope with the high-dimensionality of the resulting description space, we estimate the KL divergences in the k-th nearest neighbor (kNN) framework (instead of classical fixed size kernel methods). Query-by-example experiments show the accuracy and robustness of the method.

Robust visual tracking base on adaptively multi-feature fusion and particle filter

We present an approach for online learning of discriminative appearance models for robust multi-target tracking in a crowded scene from a single camera. Although much progress has been made in developing methods for optimal data association, there has been comparatively less work on the appearance models, which are key elements for good performance. Many previous methods either use simple features such as color histograms, or focus on the discriminability between a target and the background which does not resolve ambiguities between the different targets. We propose an algorithm for learning a discriminative appearance model for different targets. Training samples are collected online from tracklets within a time sliding window based on some spatial-temporal constraints; this allows the models to adapt to target instances. Learning uses an Ad-aBoost algorithm that combines effective image descriptors and their corresponding similarity measurements. We term the learned models as OLDAMs. Our evaluations indicate that OLDAMs have significantly higher discrimination between different targets than conventional holistic color histograms, and when integrated into a hierarchical association framework, they help improve the tracking accuracy, particularly reducing the false alarms and identity switches.