Background modelling is a critical case for background-subtraction-based approaches and also for a wide range of applications. The background generation becomes difficult when the scene is complex or an object stays for a long time in the scene. Here, the authors propose a block-based background initialisation, using the sum of absolute difference (SAD), and modelling, using a block-based entropy evaluation, with a low computational cost which making them feasible for embedded platform. In general, many background-subtraction approaches are sensitive to sudden illumination change in the scene and cannot update the background image in scenes. The proposed background modelling approach analyses the illumination change problem. The moving object detection mask is developed using a threshold selected by computing the mean of the SAD between the blocks background and the blocks of the current frame. From the qualitative and quantitative results obtained by the authors approach compared with some existing methods, the authors approach is effective for background generation and moving objects detection.
Dimensionality reduction techniques are powerful tools for face recognition, because they obtain important information from a dataset. Several dimensionality reduction methods proposed in literature have been improved thanks to preprocessing approaches. However, they also require post-processing to rectify and increase the quality of projected data. This study presents a simple and new discriminative post-processing framework to make the dimensionality reduction methods robust to outliers. In detail, the proposed approach separates features according to their scale using multidimensional ensemble empirical mode decomposition (MEEMD) and then the spatial and frequency domain processing methods are employed to preserve crucial features. The performance of the proposed method is evaluated on ORL, Extended Yale B, AR, and LFW datasets by several dimensionality reduction techniques. The experimental results demonstrate that the proposed algorithm can perform very well in face recognition.
Detecting the moving objects in a video sequence using a stationary camera is an important task for many computer vision applications. This paper proposes a background subtraction approach. As first step, the background is initialized using the block-based analysis before being updated in each incoming frame. Our background frame is generated by collecting the blocks background candidates. The block candidate selection is based on probability density function (pdf) computation. After that, we compute the absolute difference between the background frame and each frame of sequence. A noise filter is applied using the Structure/Texture decomposition in order to minimize the noise caused by background subtraction operation. The binary motion mask is formed using an adaptive threshold that was deduced from the weighted mean and variance calculation. To assure the correspondence between the current frame and the background frame, an adaptation of background model in each incoming frame is realized. After comparing results obtained from the proposed method to other existing ones, it was shown that our approach attains a higher degree of efficacy.Key Words: Motion detection, Background subtraction, Background model, Background update, Video surveillance.
IntroductionMotion detection is a paramount study field of computer vision. Its purpose is to extract the moving objects at time t in a video captured using a stationary camera. The motion detection is used for many applications. Among these applications there are video surveillance, human-machine interaction, the recognition of sign language specific to robotics applications and many others.The approaches mainly used for motion detection can be classified into three categories: the time difference methods, the analysis of optical flow, and the background subtraction methods. For the first one, the time difference being the calculation of the difference between two or more consecutive images in order to extract the moving area [1], [2], but the problem in this approach is that the detected objects are incomplete and poorly presented. The second approach is the calculation of the optical flow [3], [4] which provides all information about the movement, but the real-time implementation is difficult and calculation of
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.