Narrow Results

The aim of this paper is to propose an artificial intelligence based approach to moving object detection and tracking. Specifically, we adopt an approach to moving object detection based on self organization through artificial neural networks. Such approach allows to handle scenes containing moving backgrounds and gradual illumination variations, and achieves robust detection for different types of videos taken with stationary cameras. Moreover, for object tracking we propose a suitable conjunction between Kalman filtering, properly instanced for the problem at hand, and a matching model belonging to the class of Multiple Hypothesis Testing. To assess the validity of our approach, we experimented both proposed moving object detection and object tracking over different color video sequences that represent typical situations critical for video surveillance systems.

This paper proposes an efficient method for detecting ghost and left objects in surveillance video, which, if not identified, may lead to errors or wasted computational power in background modeling and object tracking in video surveillance systems. This method contains two main steps: the first one is to detect stationary objects, which narrows down the evaluation targets to a very small number of regions in the input image; the second step is to discriminate the candidates between ghost and left objects. For the first step, we introduce a novel stationary object detection method based on continuous object tracking and shape matching. For the second step, we propose a fast and robust inpainting method to differentiate between ghost and left objects by reconstructing the real background using the candidate's corresponding regions in the current input and background image. The effectiveness of our method has been validated by experiments over a variety of video sequences and comparisons with existing state-of-art methods.

This work presents a new approach for crowd counting and classification based upon human thermal and motion features. The technique is efficient for automatic crowd density estimation and type of motion determination. Crowd density is measured without any need for camera calibration or assumption of prior knowledge about the input videos. It does not need any human intervention so it can be used successfully in a fully automated crowd control systems. Two new features are introduced for crowd counting purpose: the first represents thermal characteristics of humans and is expressed by the ratio between their temperature and their ambient environment temperature. The second describes humans motion characteristics and is measured by the ratio between humans motion velocity and the ambient environment rigidity. Each ratio should exceed a certain predetermined threshold for human beings. These features have been investigated and proved to give accurate crowd counting performance in real time. Moreover, the two features are combined and used together for crowd classification into one of the three main types, which are: fully mobile, fully static, or mix of both types. Last but not least, the proposed system offers several advantages such as being a privacy preserving crowd counting system, reliable for homogeneous and inhomogeneous crowds, does not depend on a certain direction in motion detection, has no restriction on crowd size. The experimental results demonstrate the effectiveness of the approach.

Vital parameter monitoring systems based on video camera imagery is a growing interest field in clinical and biomedical applications. Heart rate (HR) is one of the most important vital parameters of interest in a clinical diagnostic and monitoring system. This study proposed a noncontact HR and beat length measurement system based on both motion magnification and motion detection at four different regions of interest (ROIs) (wrist, arm, neck and leg). A motion magnification based on a Chebyshev filter was utilized in order to magnify heart pulses in different ROIs that are difficult to see with the naked eye. A new measuring system based on motion detection was used to measure HR and beat length by detecting rapid motion areas in the video frame sequences that represent the heart pulses and converting video frames into a corresponding logical matrix. Video quality metrics were also used to compare our magnification system with standard Eulerian video magnification to select which one has better magnification results and gives better results for the heart pulse. The 99.3% limits of agreement between the proposed system and reference measurement fall within$\mp 1$ beats/min based on Bland and Altman test. The proposed system is expected to produce new options for further noncontact information extraction.

Ordinary projection screen is not sensitive to interaction, it cannot meet the demands of teaching, virtual reality, and other applications. Due to the fact that people always use hands to complete a variety of human–computer interaction, the finger-based interactive projection technology is worth being researched. In this paper, an ordinary monocular camera is used to acquire video frame on projection screen, and the touch signal of finger in frame is used as the input of interactive projection system. Because the differences between spatial frequency of common digital camera and the projection screen is small, the frame obtained from camera will contain moire fringe, which needs to be filtered in image frequency domain. Then the difference between current frame edge and previous frame edge is calculated to obtain moving object edge clues. According to these clues, the most possible contour curve is searched in current frame edge, and the curve is fitted by polynomial approximation method. Its curvature integration is used to match with the curvature integration of finger template curve. After that the fingers in the curve are recognized. Because color information is not needed, this method can be used to recognize gloved fingers. Finally, finger shadow is used to judge whether the finger touches projection screen to complete interactive process. The experiments of writing and collaboratively rotating picture on projector screen show that this method can effectively complete interactive operation with the projection screen and can realize the multi-user operation.

Background subtraction is a prerequisite and often the very first step employed in several high-level and real-time computer vision applications. Several parametric and non-parametric change detection algorithms employing multiple feature spaces have been proposed to date but none has proven to be robust against all challenges that can possibly be posed in a complex real-time environment. Amongst the varied challenges posed, illumination variations, shadows, dynamic backgrounds, camouflaged and bootstrapping artifacts are some of the well-known problems. This paper presents a light-weight hybrid change detection algorithm that integrates a novel combination of RGB color space and conditional YCbCr-based XCS-LBP texture descriptors (YXCS-LBP) into a modified pixel-based background model. The conditional employment of light-weight YXCS-LBP texture features with the modified Visual background extractor (ViBe) aiming at reduction in false positives, produces outperforming results without incurring much memory and computational cost. The random and time-subsampled update strategy employed with the proposed classification procedure ensures the efficient suppression of shadows and bootstrapping artifacts along with the complete retention of long-term static objects in the foreground masks. Comprehensive performance analysis of the proposed technique on publicly available Change Detection dataset (2014 CDnet dataset) demonstrates the superiority of the proposed technique over different state-of-the-art-methods against varied challenges.

Optical flow computation in dynamic image processing can be formulated as a minimization problem by a variational approach. Because solving the problem is computationally intensive, we reformulate the problem suitable for neural computing. In this paper, we propose a recurrent neural network model which may be implemented in hardware with many processing elements (neurons) operating asynchronously in parallel to achieve a possible real-time solution. We derive and prove the properties of the reformulation, as well as analyze the asymptotic stability and convergence rate of the proposed neural network. Experiments using both the test patterns and the real laboratory images are conducted.

In this paper, an algorithm for smoke region of focus (ROF) detection based on surveillance video is proposed in order to solve the problem of limited application in scenes range and imaging environment of the traditional smoke detection algorithm. The frog vision perception model is used in this algorithm. First the suspect regions are detected, and then the static and the dynamic features of the regions are chosen for the smoke ROF detection. Experimental results show that the algorithm is efficient and significant for improving the operational rate of the detection.

The aim of this study is to remotely measure cardiac activity (heart pulse, total cycle length and pulse width) from videos based on a head motion at different positions of the head (front, back and side). As the head motion resulting from the cardiac cycle of blood from the heart to the head via the carotid arteries is not visible to the naked eye and to preserve the signal strength in the video, we used wavelet decomposition and a Chebychev filter to develop a standard Eulerian video magnification in terms of noise removal and execution time. We used both magnification systems to measure cardiac activity and statistically compare the results using Bland–Altman method. Also, we proposed a new video quality system based on fuzzy interface system to select which magnification system has better magnification quality and gives better results for the heart pulse rate. The experimental results on several videos captured from 10 healthy subjects show that the proposed contactless system of heart pulse has an accuracy of 98.3% when magnified video based on the developing magnification system was used and an accuracy of 97.4% when magnified video based on Eulerian magnification system was used instead. The proposed system has low computational complexity, making it suitable for advancing health care applications, mobile health applications and telemedicine.

Motion detection and object tracking play important roles in unsupervised human–machine interaction systems. Nevertheless, the human–machine interaction would become invalid when the system fails to detect the scene objects correctly due to occlusion and limited field of view. Thus, robust long-term tracking of scene objects is vital. In this paper, we present a 3D motion detection and long-term tracking system with simultaneous 3D reconstruction of dynamic objects. In order to achieve the high precision motion detection, an optimization framework with a novel motion pose estimation energy function is provided in the proposed method by which the 3D motion pose of each object can be estimated independently. We also develop an accurate object-tracking method which combines 2D visual information and depth. We incorporate a novel boundary-optimization segmentation based on 2D visual information and depth to improve the robustness of tracking significantly. Besides, we also introduce a new fusion and updating strategy in the 3D reconstruction process. This strategy brings higher robustness to 3D motion detection. Experiments results show that, for synthetic sequences, the root-mean-square error (RMSE) of our system is much smaller than Co-Fusion (CF); our system performs extremely well in 3D motion detection accuracy. In the case of occlusion or out-of-view on real scene data, CF will suffer the loss of tracking or object-label changing, by contrast, our system can always keep the robust tracking and maintain the correct labels for each dynamic object. Therefore, our system is robust to occlusion and out-of-view application scenarios.

Intelligent surveillance aims at conceiving reliable and efficient systems that are able to detect and track moving objects in complicated real world scenes. This paper proposes an innovative 3D stationary wavelet-based motion detection technique that fuses spatial and temporal analysis in a single 3D transform. This single transform is composed of applying a 2D transform in the spatial domain followed by 1D transform in the time domain. The results of the proposed technique are compared favorably with those of the recently used stationary wavelet-based technique. In addition of being accurate and has reasonable complexity of O(N²log N), the proposed technique is robust to real world scene variations, including nonuniform and time-varying illumination.

With the progress of microelectronic devices, more attention has been paid to carbon nanotube (CNTs) because of their excellent mechanical and electrical properties. A microfluidic-based method for fabricating oriented CNT strain-sensitive fibers is proposed in this work. By manipulating CNTs, NaAlg and CaCl₂ solutions in a three-coaxial laminar flow glass capillary device, CNTs can be arranged in an oriented manner and solidified into microfibers with a hydrogel shell through a chemical polymerization reaction. The diameter of the fiber could be controlled by the microfluidic device. Scanning electron microscopy and Raman spectroscopy show that the CNTs are perfectly aligned along the fiber axis under the influence of the viscous drag force of the fluid, and the electrical and mechanical properties of the CNT fibers are obviously strengthened. Experiments suggest that the devices based on the oriented CNT–alginate microfibers are able to achieve simple motion monitoring function. The proposed microfluidic method is simple, cost effective and can be applied to produce functional nanomaterial fibers for application in flexible devices.

This paper presents a mobile object detection algorithm which performs with two consecutive stereo images. Like most motion detection methods, the proposed one is based on dense stereo matching and optical flow (OF) estimation. Noting that the main computational cost of existing methods is related to the estimation of OF, we propose to use a fast algorithm based on Lucas–Kanade paradigm. We then derive a comprehensive uncertainty model by taking into account all the estimation errors occurring during the process. In contrast with most previous works, we rigorously expand the error related to vision based ego-motion estimation. Finally, we present a comparative study of performance on the challenging KITTI dataset which demonstrates the effectiveness of the proposed approach.

Amongst the motion detection and correction algorithms during the scanning procedures, data-processing methods are the most frequently proposed solution to detect and correct patient motions. There are different distance metrics which have been used to detect the patient motions using information contained in the projections. Unfortunately, the performance of usually used metrics is low in the case of small motions while detecting the motions with magnitude of 1 pixel and smaller are very important in the accuracy of diagnosis. In this work, a new distance metric, normalized prediction of projection data algorithm (NPPDA) is developed based on the linear prediction filter. The performance of the NPPDA is quantitatively evaluated and compared with usual distance metrics by different experimental studies. A high detection rate is achieved by means of the newly developed distance metric, NPPDA.

In this paper, a method for detecting the nucleus movement of oocytes during the enucleation process based on the mean drift algorithm is proposed, including the following steps: 1. Establish the target model ROI_ini and calculate the probability density histogram; 2. Establish the target candidate model ROI_candi and calculate the probability density histogram; 3. Use the Bhattacharyya coefficient to compare the similarity of the target model and the target candidate model; 4. Locate the moving target. This method is a universal nucleus motility detection method, which improves the limitations of the traditional mean drift target tracking algorithm, solves the problem of nucleus motion detection under the conditions of low microscopic image resolution, change in nucleus shape during enucleation, and large differences in the shape of different oocyte nuclei. This method can be used in the field of somatic cell nuclear transplantation, which can greatly improve the accuracy of oocyte enucleation, reduce cell damage, and further improve the development potential of recombinant cells.

Optical flow computation in dynamic image processing can be formulated as a minimization problem by a variational approach. Because solving the problem is computationally intensive, we reformulate the problem suitable for neural computing. In this paper, we propose a recurrent neural network model which may be implemented in hardware with many processing elements (neurons) operating asynchronously in parallel to achieve a possible real-time solution. We derive and prove the properties of the reformulation, as well as analyze the asymptotic stability and convergence rate of the proposed neural network. Experiments using both the test patterns and the real laboratory images are conducted.

We propose simple analog MOS circuits with a correlation model based on the insect motion detectors aiming at the realization of fundamental motion-sensing systems. The model makes the circuit structure quite simple, compared with conventional velocity sensing circuits. SPICE simulation results indicate that the proposed circuits compute local velocities of the moving light spot and have direction selectivity for the spot, which implies that a high-resolution motion-sensing chip can be realized by current analog VLSI technology.