Mean-shift analysis is a general nonparametric clustering technique based on density estimation for the analysis of complex feature spaces. The algorithm consists of a simple iterative procedure that shifts each of the feature points to the nearest stationary point along the gradient directions of the estimated density function. It has been successfully applied to many applications such as segm...

In most low-level computer vision problems, very little information (if any) is known about the true underlying probability density function, such as its shape, number of mixture components, etc.. Due to this lack of knowledge, parametric approaches are less relevant, rather one has to rely on non-parametric methods. In this note we consider the construction and convergence proof of the non-par...

An object tracking algorithm using the Mean Shift framework is presented which is largely invariant to both partial and full occlusions, complex backgrounds and change in scale. Multiple features are used to gain a descriptive representation of the target object. Image moments are used to determine the scale of the target object. A kalman filter is used to successfully track the target object t...

The mean shift algorithm is a powerful clustering technique, which is based on an iterative scheme to detect modes in a probability density function. It has been utilized for image segmentation by seeking the modes in a feature space composed of spatial and color information. Although the modes of the feature space can be efficiently calculated in that scheme, different optimization techniques ...

In this paper, we present an improved mean shift for robust object tracking in complex environment. Traditional RGB color model used in mean shift tracker is sensitive to interference from similar background. In order to solve this problem, a new saliency-color target model is proposed through using the state-of-the-art target representation and updated background-weighed method. In addition, t...

In this work, boosting the efficiency of Mean-Shift Tracking using random sampling is proposed. We obtained the surprising result that mean-shift tracking requires only very few samples. Our experiments demonstrate that robust tracking can be achieved with as few as even 5 random samples from the image of the object. As the computational complexity is considerably reduced and becomes independen...

We propose a novel Mean-Shift method for data clustering, called Robust (RMS). A new update equation point iterates is proposed, mixing the ones of standard (MS) and Blurring (BMS). Despite its simplicity, proposed has not been studied so far. RMS can be set up in both kernel-based nearest-neighbor (NN)-based fashion. Since rule closer to BMS, convergence conjectured based on Chen’s BMS theorem...