基于AdaBoost的公交客流量统计算法

为解决复杂场景目标识别中伪目标的干扰问题,采用基于AdaBoost分类的方法分析疑似目标的三维轨迹,结合真实目标的共有的特征信息,进一步分类真实目标与伪目标。首先,根据深度相机获取的深度图像提取疑似目标的人头区域,利用Kalman滤波跟踪得到的二维轨迹。其次,通过摄像机标定将目标的二维轨迹转换为空间中的三维轨迹。最后,利用AdaBoost训练正负样本得到强分类器,进一步分类真实目标与伪目标。实验结果表明,该方法能够有效的提高目标识别的精度,对复杂场景下的目标识别具有良好的适应性。     

基于三维模型的粒子滤波行人跟踪算法

针对传统行人跟踪算法得到运动轨迹与真实轨迹差异巨大的问题,提出一种基于三维模型的粒子滤波行人跟踪算法.该方法利用摄像机标定信息和图像帧信息建立行人的三维模型,解决图像中目标尺度的变化问题,并得到目标的真实运动轨迹.同时该方法利用双指数预测模型对粒子滤波算法进行优化,以解决短时遮挡问题,同时降低运算复杂度.实验表明,基于三维模型的粒子滤波行人跟踪算法能够较准确地建立行人三维模型,对比标准粒子滤波和KPF算法,能够对行人进行有效跟踪,对短时遮挡和尺度变化有较强的鲁棒性.     

Model-driven active visual tracking

We have previously demonstrated that the performance of tracking algorithms can be improved by integrating information from multiple cues in a model-driven Bayesian reasoning framework. Here we extend our work to active vision tracking, with variable camera geometry. Many existent active tracking algorithms avoid the problem of variable camera geometry by tracking view independent features, such as corners and lines. However, the performance of algorithms based on those single features will greatly deteriorate in the presence of specularities and dense clutter. We show, by integrating multiple cues and updating the camera geometry on-line, that it is possible to track a complicated object moving arbitrarily in three-dimensional (3D) space.We use a four degree-of-freedom (4-DoF) binocular camera rig to track three focus features of an industrial object, whose complete model is known. The camera geometry is updated by using the rig control commands and kinematic model of the stereo head. The extrinsic parameters are further refined by interpolation from a previously sampled calibration of the head work space.The 2D target position estimates are obtained by a combination of blob detection, edge searching and gray-level matching, with the aid of model geometrical structure projection using current estimates of camera geometry. The information is represented in the form of a probability density distribution, and propagated in a Bayes Net. The Bayesian reasoning that is performed in the 2D image is coupled by the rigid model geometry constraint in 3D space.An αβ filter is used to smooth the tracking pursuit and to predict the position of the object in the next iteration of data acquisition. The solution of the inverse kinematic problem at the predicted position is used to control the position of the stereo head.Finally, experiments show that the target undertaking arbitrarily 3D motion can be successfully tracked in the presence of specularities and dense clutter.     

Manifold surface reconstruction of an environment from sparse Structure-from-Motion data

The majority of methods for the automatic surface reconstruction of an environment from an image sequence have two steps: Structure-from-Motion and dense stereo. From the computational standpoint, it would be interesting to avoid dense stereo and to generate a surface directly from the sparse cloud of 3D points and their visibility information provided by Structure-from-Motion. The previous attempts to solve this problem are currently very limited: the surface is non-manifold or has zero genus, the experiments are done on small scenes or objects using a few dozens of images. Our solution does not have these limitations. Furthermore, we experiment with hand-held or helmet-held catadioptric cameras moving in a city and generate 3D models such that the camera trajectory can be longer than one kilometer.     

Navidget for 3D interaction: Camera positioning and further uses

This paper presents an extended version of Navidget. Navidget is a new interaction technique for camera positioning in 3D environments. This technique derives from the point-of-interest (POI) approaches where the endpoint of a trajectory is selected for smooth camera motions. Unlike the existing POI techniques, Navidget does not attempt to automatically estimate where and how the user wants to move. Instead, it provides good feedback and control for fast and easy interactive camera positioning. Navidget can also be useful for distant inspection when used with a preview window. This new 3D user interface is totally based on 2D inputs. As a result, it is appropriate for a wide variety of visualization systems, from small handheld devices to large interactive displays. A user study on TabletPC shows that the usability of Navidget is very good for both expert and novice users. This new technique is more appropriate than the conventional 3D viewer interfaces in numerous 3D camera positioning tasks. Apart from these tasks, the Navidget approach can be useful for further purposes such as collaborative work and animation.     

Human fall detection using normalized shape aspect ratio

In video surveillance, automatic human fall detection is important to protect vulnerable groups such as the elderly. When the camera layout varies, the shape aspect ratio (SAR) of a human body may change substantially. In order to rectify these changes, in this paper, we propose an automatic human fall detection method using the normalized shape aspect ratio (NSAR). A calibration process and bicubic interpolation are implemented to generate the NSAR table for each camera. Compared with some representative fall detection methods using the SAR, the proposed method integrates the NSAR with the moving speed and direction information to robustly detect human fall, as well as being able to detect falls toward eight different directions for multiple humans. Moreover, while most of the existing fall detection methods were designed only for indoor environment, experimental results demonstrate that this newly proposed method can effectively detect human fall in both indoor and outdoor environments.

     

Map Building Fusing Acoustic and Visual Information using Autonomous Underwater Vehicles

We present a system for automatically building three‐dimensional (3‐D) maps of underwater terrain fusing visual data from a single camera with range data from multibeam sonar. The six‐degree‐of‐freedom location of the camera relative to the navigation frame is derived as part of the mapping process, as are the attitude offsets of the multibeam head and the onboard velocity sensor. The system uses pose graph optimization and the square root information smoothing and mapping framework to simultaneously solve for the robot's trajectory, the map, and the camera location in the robot's frame. Matched visual features are treated within the pose graph as images of 3‐D landmarks, while multibeam bathymetry submap matches are used to impose relative pose constraints linking robot poses from distinct tracklines of the dive trajectory. The navigation and mapping system presented works under a variety of deployment scenarios on robots with diverse sensor suites. The results of using the system to map the structure and the appearance of a section of coral reef are presented using data acquired by the Seabed autonomous underwater vehicle.     

Visual dynamic model based on self-organizing maps for supervision and fault detection in industrial processes

Visual data mining techniques have experienced a growing interest for processing and interpretation of the large amounts of multidimensional data available in current industrial processes. One of the approaches to visualize data is based on self-organizing maps (SOM), which define a projection of the input space onto a 2D or 3D space that can be used to obtain visual representations. Although these techniques have been usually applied to visualize static relations among the process variables, they have proven to be very useful to display dynamic features of the processes. In this work, an approach based on the SOM to model the dynamics of multivariable processes is presented. The proposed method identifies the process conditions (clusters) and the probabilities of transition among them, using the trajectory followed by the input data on the 2D visualization space. Furthermore, a new method of residual computation for fault detection and identification that uses the dynamic information provided by the model of transitions is proposed. The proposed method for modeling and fault identification has been applied to supervise a real industrial plant and the results are included.     

3D Face Modeling from Perspective-Views and Contour-Based Generic-Model

Three-dimensional human head modeling is useful in video-conferencing or other virtual reality applications. However, manual construction of 3D models using CAD tools is often expensive and time-consuming. Here we present a robust and efficient method for the construction of a 3D human head model from perspective images viewing from different angles. In our system, a generic head model is first used, then three images of the head are required to adjust the deformable contours on the generic model to make it closer to the target head. Our contributions are as follows. Our system uses perspective images that are more realistic than orthographic projection approximations used in earlier works. Also, for shaping and positioning face organs, we present a method for estimating the camera focal length and the 3D coordinates of facial landmarks when the camera transformation is known. We also provide an alternative for the 3D coordinates estimation using epipolar geometry when the extrinsic parameters are absent. Our experiments demonstrate that our approach produces good and realistic results.     

Viewpoint image generation for head tracking 3D display using multi‐camera and approximate depth information

A simple and high image quality method for viewpoint image synthesis from multi‐camera images for a stereoscopic 3D display using head tracking is proposed. In this method, slices of images for depth layers are made using approximate depth information, the slices are linearly blended corresponding to the distance between the viewpoint and cameras at each layer, and the layers are overlaid from the perspective of viewpoint. Because the linear blending automatically compensates for depth error because of the visual effects of depth‐fused 3D (DFD), the resulting image is natural to observer's perception. Smooth motion parallax of wide depth range objects induced by viewpoint movement for left‐and‐right and front‐and‐back directions is achieved using multi‐camera images and approximate depth information. Because the calculation algorithm is very simple, it is suitable for real time 3D display applications.     

Robot visual servoing with iterative learning control

This paper presents an iterative learning scheme for vision-guided robot trajectory tracking. First, a stability criterion for designing iterative learning controller is proposed. It can be used for a system with initial resetting error. By using the criterion, one can convert the design problem into finding a positive definite discrete matrix kernel and a more general form of learning control can be obtained. Then, a three-dimensional (3D) trajectory tracking system with a single static camera to realize robot movement imitation is presented based on this criterion.     

A theoretical 2D image model for locating 3D targets

To construct a water quality monitoring system, challenging issues need to be addressed regarding the acquisition of target information (e.g. 3D location and occlusion) as well as the behavioural analysis of aquatic organisms. This paper presents a novel 3D information acquisition and location method, by means of an information acquisition platform consisting of a monitoring terminal, frame grabbers, a single camera and a single mirror. Using this platform, we propose a theoretical 2D image model for locating 3D targets and then validate it using data obtained from both real and artificial fish. The proposed model is based on the principles of light refraction, plane mirror imaging, underwater objects and camera imaging as well as the technologies of digital to analog conversion and object segmentation. In contrast with existing methods, our method can accurately reflect 3D information of aquatic organisms, thus providing critical technical support for the development of water quality monitoring systems in the future.     

Trip Synopsis: 60km in 60sec

Computerized route planning tools are widely used today by travelers all around the globe, while 3D terrain and urban models are becoming increasingly elaborate and abundant. This makes it feasible to generate a virtual 3D flyby along a planned route. Such a flyby may be useful, either as a preview of the trip, or as an after‐the‐fact visual summary. However, a naively generated preview is likely to contain many boring portions, while skipping too quickly over areas worthy of attention. In this paper, we introduce 3D trip synopsis: a continuous visual summary of a trip that attempts to maximize the total amount of visual interest seen by the camera. The main challenge is to generate a synopsis of a prescribed short duration, while ensuring a visually smooth camera motion. Using an application‐specific visual interest metric, we measure the visual interest at a set of viewpoints along an initial camera path, and maximize the amount of visual interest seen in the synopsis by varying the speed along the route. A new camera path is then computed using optimization to simultaneously satisfy requirements, such as smoothness, focus and distance to the route. The process is repeated until convergence. The main technical contribution of this work is a new camera control method, which iteratively adjusts the camera trajectory and determines all of the camera trajectory parameters, including the camera position, altitude, heading, and tilt. Our results demonstrate the effectiveness of our trip synopses, compared to a number of alternatives.     

Virtual 3D City Model for Navigation in Urban Areas

In this paper, we propose to study the integration of a new source of a priori information, which is the virtual 3D city model. We study this integration for two tasks: vehicles geo-localization and obstacles detection. A virtual 3D city model is a realistic representation of the evolution environment of a vehicle. It is a database of geographical and textured 3D data. We describe an ego-localization method that combines measurements of a GPS (Global Positioning System) receiver, odometers, a gyrometer, a video camera and a virtual 3D city model. GPS is often consider as the main sensor for localization of vehicles. But, in urban areas, GPS is not precise or even can be unavailable. So, GPS data are fused with odometers and gyrometer measurements using an Unscented Kalman Filter (UKF). However, during long GPS unavailability, localization with only odometers and gyrometer drift. Thus, we propose a new observation of the location of the vehicle. This observation is based on the matching between the current image acquired by an on-board camera and the virtual 3D city model of the environment. We also propose an obstacle detection method based on the comparison between the image acquired by the on-board camera and the image extracted from the 3D model. The following principle is used: the image acquired by the on-board camera contains the possible dynamic obstacles whereas they are absent from the 3D model. The two proposed concepts are tested on real data.     

Calibration of an active binocular head

In this paper, we show how an active binocular head, the IIS head, can be easily calibrated with very high accuracy. Our calibration method can also be applied to many other binocular heads. In addition to the proposal and demonstration of a four-stage calibration process, there are three major contributions in this paper. First, we propose a motorized-focus lens (MFL) camera model which assumes constant nominal extrinsic parameters. The advantage of having constant extrinsic parameters is to having a simple head/eye relation. Second, a calibration method for the MFL camera model is proposed in this paper, which separates estimation of the image center and effective focal length from estimation of the camera orientation and position. This separation has been proved to be crucial; otherwise, estimates of camera parameters would be very noise-sensitive. Thirdly, we show that, once the parameters of the MFL camera model is calibrated, a nonlinear recursive least-square estimator can be used to refine all the 35 kinematic parameters. Real experiments have shown that the proposed method can achieve accuracy of one pixel prediction error and 0.2 pixel epipolar error, even when all the joints, including the left and right focus motors, are moved simultaneously. This accuracy is good enough for many 3D vision applications, such as navigation, object tracking and reconstruction     

Multiple human tracking in high-density crowds

In this paper, we introduce a fully automatic algorithm to detect and track multiple humans in high-density crowds in the presence of extreme occlusion. Typical approaches such as background modeling and body part-based pedestrian detection fail when most of the scene is in motion and most body parts of most of the pedestrians are occluded. To overcome this problem, we integrate human detection and tracking into a single framework and introduce a confirmation-by-classification method for tracking that associates detections with tracks, tracks humans through occlusions, and eliminates false positive tracks. We use a Viola and Jones AdaBoost detection cascade, a particle filter for tracking, and color histograms for appearance modeling. To further reduce false detections due to dense features and shadows, we introduce a method for estimation and utilization of a 3D head plane that reduces false positives while preserving high detection rates. The algorithm learns the head plane from observations of human heads incrementally, without any a priori extrinsic camera calibration information, and only begins to utilize the head plane once confidence in the parameter estimates is sufficiently high. In an experimental evaluation, we show that confirmation-by-classification and head plane estimation together enable the construction of an excellent pedestrian tracker for dense crowds.     

Stable real-time 3D tracking using online and offline information

We propose an efficient real-time solution for tracking rigid objects in 3D using a single camera that can handle large camera displacements, drastic aspect changes, and partial occlusions. While commercial products are already available for offline camera registration, robust online tracking remains an open issue because many real-time algorithms described in the literature still lack robustness and are prone to drift and jitter. To address these problems, we have formulated the tracking problem in terms of local bundle adjustment and have developed a method for establishing image correspondences that can equally well handle short and wide-baseline matching. We then can merge the information from preceding frames with that provided by a very limited number of keyframes created during a training stage, which results in a real-time tracker that does not jitter or drift and can deal with significant aspect changes.     

3D measures exploitation for a monocular semi-supervised fall detection system

Falls have been reported as the leading cause of injury-related visits to emergency departments and the primary etiology of accidental deaths in elderly. Thus, the development of robust home surveillance systems is of great importance. In this article, such a system is presented, which tries to address the fall detection problem through visual cues. The proposed methodology utilizes a fast, real-time background subtraction algorithm, based on motion information in the scene and pixels intensity, capable to operate properly in dynamically changing visual conditions, in order to detect the foreground object. At the same time, it exploits 3D space’s measures, through automatic camera calibration, to increase the robustness of fall detection algorithm which is based on semi-supervised learning approach. The above system uses a single monocular camera and is characterized by minimal computational cost and memory requirements that make it suitable for real-time large scale implementations.