Budi Sugandi

Tracking of Moving Objects by Using a Low Resolution Image

Automatic detection and tracking of moving object is very important task for security system, monitoring activity and surveillance application. It is, however, still under the developmental stage and needs to be robust performance when applied in an unconstrained environment. Many approaches have been developed to detect motion, namely, optical flow, frame difference, background subtraction and skin color extraction. These methods are sensitive to illumination changes and small movement in the background such as moving leaves of trees that will cause inaccurate detection. Some techniques have been applied to reduce this kind of noise. In this paper we propose a new method by using a low resolution image to reduce that kind of noise so we can get an accurate tracking object.

A color-based particle filter for multiple object tracking in an outdoor environment

Tracking multiple objects is more challenging than tracking a single object. Some problems arise in multiple-object tracking that do not exist in single-object tracking, such as object occlusion, the appearance of a new object and the disappearance of an existing object, updating the occluded object, etc. In this article, we present an approach to handling multiple-object tracking in the presence of occlusions, background clutter, and changing appearance. The occlusion is handled by considering the predicted trajectories of the objects based on a dynamic model and likelihood measures. We also propose target-model-update conditions, ensuring the proper tracking of multiple objects. The proposed method is implemented in a probabilistic framework such as a particle filter in conjunction with a color feature. The particle filter has proven very successful for nonlinear and non-Gaussian estimation problems. It approximates a posterior probability density of the state, such as the object’s position, by using samples or particles, where each state is denoted as the hypothetical state of the tracked object and its weight. The observation likelihood of the objects is modeled based on a color histogram. The sample weight is measured based on the Bhattacharya coefficient, which measures the similarity between each sample’s histogram and a specified target model. The algorithm can successfully track multiple objects in the presence of occlusion and noise. Experimental results show the effectiveness of our method in tracking multiple objects.

A moving object tracking based on color information employing a particle filter algorithm

In this article, we present a new algorithm to track a moving object based on color information employing a particle filter algorithm. Recently, a particle filter has been proven very successful for nonlinear and non-Gaussian estimation problems. It approximates a posterior probability density of the state, such as the object position, by using samples which are called particles. The probability distribution of the state of the tracked object is approximated by a set of particles, where each state is denoted as the hypothetical state of the tracked object and its weight. The particles are propagated according to a state space model. Here, the state is treated as the position of the object. The weight is considered as the likelihood of each particle. For this likelihood, we consider the similarity between the color histogram of the tracked object and the region around the position of each particle. The Bhattacharya distance is used to measure this similarity. Finally, the mean state of the particles is treated as the estimated position of the object. Experiments were performed to confirm the effectiveness of this method to track a moving object.

A block matching technique for object tracking employing peripheral increment sign correlation image

In this paper, we propose a new technique to detect and track the interested moving person employing a block matching technique. The blocks are defined by dividing the image frame into non-overlapping square parts and made in the previous and current frame. The blocks are made based on the peripheral increment sign correlation image. The increment sign of each block is calculated by comparing the pixel value of the neighborhood to the considered pixel. The correlation value between each block in the previous and current frame is calculated to evaluate the matching block. The high correlation value shows that the block is matched. The interested moving object is determined when the number of matching blocks in the previous and current frame is higher than a certain threshold value. Our method is implemented in real time application by using a camera and the satisfied results are achieved.

Tracking of multiple moving objects under outdoor environment using color-based particle filter

Object tracking is a challenging problem due to the presence of noise, occlusion, clutter and dynamic change in the scene other than the motion of object of interest. A variety of tracking algorithms has been proposed and implemented to overcome these difficulties, but there are still some problems need to be covered. This paper proposed an algorithm to track the moving objects employing a color-based particle filter considering the single object and multiple objects in outdoor environment. We rely on Bhattacharya distance to measure the similarity between the color distribution of the target model and particles. We propose also a target-model update condition, ensuring the proper tracking object. The method is capable to track successfully the object in different outdoor environment in the presence of occlusion, clutter and dynamic change. Some experimental results and data show the feasibility and the effectiveness of our method.

Object Tracking Based on Color Information Employing Particle Filter Algorithm

The increasing interest in the object tracking is motivated by a huge number of promising applications that can now be tackled in real-time applications. These applications include performance analysis, surveillance, video-indexing, smart interfaces, teleconferencing and video compression and so on. However, object tracking can be extremely complex and timeconsuming especially when it is done in outdoor environments. Here, we can mention some problems of object tracking in outdoor environments such as fake-motion background, illumination changes, shadows and presence of clutter. A variety of tracking algorithms have been proposed and implemented to overcome these difficulties. They can be roughly classified into two categories: deterministic methods and stochastic methods. Deterministic methods typically track the object by performing an iterative search for a similarity between the template image and the current one. The algorithms which utilize the deterministic method are background subtraction (Heikkila & Silven, 1999; Stauffer & Grimson, 1999; McIvor, 2000; Liu et al., 2001), inter-frame difference (Lipton et al., 1998; Collins et al., 2000), optical flow (Meyer et al., 1998), skin color extraction (Cho et al., 2001; Phung et al., 2003) and so on. On the other hand, the stochastic methods use the state space to model the underlying dynamics of the tracking system such as Kalman filter (Broida & Chellappa, 1986; Arulampalam et al., 2002) and particle filter (Isard & Black, 1998; Kitagawa, 1996; Gordon et al., 1993; Ristic et al., 2004). Kalman filter is a common approach for dealing with target tracking in the probabilistic framework. In a linear-Gaussian model with linear measurement, there is always only one mode in the posterior probability density function (pdf). The Kalman filter can be used to propagate and update the mean and covariance of the distribution of this model (Arulampalam et al., 2002). But it cannot resolve the tracking problem when the model is nonlinear and non-Gaussian (Tanizaki, 1987). For nonlinear or non-Gaussian problems, it is impossible to evaluate the distributions analytically and many algorithms have been proposed to approximate them. The extended Kalman filter can deal to this problem, but still has a problem when the nonlinearity and non-Gaussian cannot be approximated accurately. To overcome those problems, particle filter has been introduced by many researchers and become popular algorithm to estimate the problem of nonlinear and non-Gaussian estimation framework. The particle filter, also known as sequential Monte Carlo (Kitagawa, 1996), is the most popular approach which recursively constructs the posterior pdf of the state space using Monte Carlo integration. It approximates a posterior probability density of the state such as the object position by using samples or

Automatic Detection and Tracking of Moving Object Employing a Particle Filter

We proposed a method for automatic detection and tracking of moving object employing a particle filter in conjunction with a color feature method. The particle filtering is used because it is robust for non-linear and non-Gaussian dynamic state estimation problems and performs well when clutter and occlusions are present on the image. A histogram-based framework is used to describe the color feature of the target object. Bhattacharyya distance is used to measure the similarity between each samples histogram with a specified target model. The target model update is performed to obtain the best match to the target model. The method is capable to detect and track successfully the moving object in different outdoor environment based on variance of the samples and an appearance condition. The experimental results and data show the feasibility and the effectiveness of our method.

A COLOR FEATURES‐BASED METHOD FOR OBJECT TRACKING EMPLOYING A PARTICLE FILTER ALGORITHM

We proposed a method for object tracking employing a particle filter based on color feature method. A histogram‐based framework is used to describe the features. Histograms are useful because they have property that they allow changes in the object appearance while the histograms remain the same. Particle filtering is used because it is very robust for non‐linear and non‐Gaussian dynamic state estimation problems and performs well when clutter and occlusions are present on the image. Bhattacharyya distance is used to weight the samples in the particle filter by comparing each sample’s histogram with a specified target model and it makes the measurement matching and sample’s weight updating more reasonable. The method is capable to track successfully the moving object in different outdoor environment with and without initial positions information, and also, capable to track the moving object in the presence of occlusion using an appearance condition. In this paper, we propose a color features‐based method ...

Face direction estimation based on eigenspace technique

In this paper, we propose a method for identification of persons using feature points which are taken from different angle and estimating of face direction using eigenspace technique. The face images are extracted on image sequences which are captured by a camera using image processing techniques. The face directions are estimated based on the eigenspace technique. Furthermore, the locations of face feature points are determined using a separability filter. Finally, the human faces are identified using the statistical features. The feature points for face identification are six points which are obtained the length of eyes, nostrils and angle of mouth. The experiments are performed on real time video image. The satisfactory results are shown along with discussions.