Search

Detection capability of space camera is regarded as the principal parameter in the design of space camera. On the basis of the concept of signal-to-noise ratio (SNR) and through the research of principles and derivation of theories, detection capabilities under static condition and the condition of relative motion were analyzed and compared. With a set of typical parameters given, the change of detectable magnitude with the changes of parameters was investigated. Considering the real operating environment of space camera, the necessity of temperature control in the design of space camera was discussed. The theoretical analysis provides a reference for the design of space camera.

This paper proposes a novel method via dynamic tree to solve the non-rigid registration of point sets with large shape difference which is a difficult problem for existing methods. Affine ICP algorithm with bidirectional distance is employed to evaluate the similarity between two point sets, and then non-rigid registration is conducted on similar models and subjects. Subjects with accurate registration results are added in the dynamic tree. These steps are repeated until all subjects are linked in the tree. Therefore, large shape difference is divided into several small deformations by the intermediate point sets, and the registration results of model and every subject in the tree are satisfactory. Experimental results show that our method incredibly improves the accuracy of the registration with two point sets bearing large shape difference compared with existing approaches.

This paper proposes a new probability iterative closest point approach with bounded scale based on expectation maximization (EM) estimation for scaling registration of point sets with noise. The bounded scale ICP algorithm can handle the case with different scales, but it could not effectively yield the alignment of point sets with noise. Aiming at improving the registration precision, a Gaussian probability model is integrated into the bounded scale registration. The proposed method can be solved by the E-step and M-step. In the E-step, we can build up the one-to-one correspondence between two point sets. In the M-step, the scale transformation which consists of the rotation matrix, translation vector, and the scale factor is solved by singular value decomposition (SVD) method and the properties of parabola. Then, the Gaussian model is updated via the distance and variance between the transformed point sets. As one-to-one correspondence is adopted for the scaling registration of point sets with noise, the proposed method improves the performance significantly with high precision and fast speed. Experimental results demonstrate that the proposed algorithm is more accurate and fast.

The article presents a new method to detect small moving targets in space surveillance. Image sequences are processed to detect and track targets under the assumption that the data samples are spatially registered. Maximum value projection and normalization are performed to reduce the data samples and eliminate the background clutter. Targets are then detected through connected component analysis. The velocities of the targets are estimated by centroid localization and least squares regression. The estimated velocities are utilized to track the targets. A sliding neighborhood operation is performed prior to target detection to significantly reduce the computation while preserving as much target information as possible. Actual data samples are acquired to test the proposed method. Experimental results show that the method can efficiently detect small moving targets and track their traces accurately. The centroid locating precision and tracking accuracy of the method are within a pixel.

To satisfy the requirement of the astronomical observation, a novel timing sequence of frame transfer CCD is proposed. The multiple functions such as the adjustments of work pattern, exposure time and frame frequency are achieved. There are four work patterns: normal, standby, zero exposure and test. The adjustment of exposure time can set multiple exposure time according to the astronomical observation. The fame frequency can be adjusted when dark target is imaged and the maximum exposure time cannot satisfy the requirement. On the design of the video processing, offset correction and adjustment of multiple gains are proposed. Offset correction is used for eliminating the fixed pattern noise of CCD. Three gains pattern can improve the signal to noise ratio of astronomical observation. Finally, the images in different situations are collected and the system readout noise is calculated. The calculation results show that the designs in this paper are practicable.

This paper proposes a trajectory generation method for mobile robot by extending the trajectory segment iteratively. Considering the practical uses of the mobile robot like service purpose in the real world, it requires that the trajectory planning is done adaptively by depending on the situation changes. For that, designated waypoints for future motion are given sequentially from the environment recognition process and thus trajectory generation should be processed in dynamical situation. In this paper, we described an iterative extension of the polynomial-based trajectory by considering the designated waypoint, the continuity with the existing trajectory segment and the stoppage constraint at the end of each trajectory segment. Proposed method provides the next trajectory segment for a certain period by referring to these conditions. Resultantly, iteratively generated trajectory segments connect each other smoothly and organize a single continuous trajectory. Computer simulations with proposed method were conducted and the results were shown. They showed proposed method realizes trajectory generation based on sequential extension of the trajectory segment.

The Vector Field Histogram (VFH) is a classical motion planning algorithm which is widely used to handle the trajectory planning problem of mobile robots. However, the traditional VFH algorithm is rarely applied to autonomous vehicles due to the vehicle's well-known non-holonomic constraints, especially in urban environments. To address this problem, we propose a constrained VFH algorithm which takes both kinematic and dynamic constraints of the vehicle into consideration. The goal is achieved via two contributions that concern both kinematic and dynamic constraints of the vehicle. First, we develop a new active region for VFH to guarantee that all states within the region are reachable for the vehicle. Second, we improve the cost function to guide the search to favor feasible motion direction for the vehicle. The proposed algorithm is extensively tested in various simulated urban environments, and experimental results validate its efficiency.

In the traditional trajectory planning methods, a feasible, collision-free trajectory is generated to guide the vehicle. But generally the vehicle cannot follow the trajectory without tracking deviation because of the vehicle kinematical constraints and the performance of control algorithm. In this paper, State-Statistical Model (SSM) based trajectory-band planning method is proposed to predict the vehicle motion during the vehicle tracks the trajectory. In this method, the statistics of historical states are used to build the SSM which is a normal distribution model of tracking deviation in different segments of curvature radius and velocity. According to the SSM, the inaccessible states of vehicle can be obtained to search the best trajectory and the tracking deviation boundary can be calculated on the trajectory. Then the best trajectory is used as the base line to generate the trajectory-band of which the halfband width is the deviation boundary value. As a result, the trajectory-band can represent the maximum range of vehicle motion accurately.

This paper proposes a trajectory generation method for mobile robots based on trajectory segment extension. The trajectory for mobile robot must be generated by depending on the conditions during the robot traveling. Such conditions to design the trajectory are given when the robot comes around those locations and are not provided before arriving around there. From this point of view, we have started to develop an on-line trajectory generation method based on extending the trajectory segment that is the low-order polynomial curve. We already proposed a trajectory extension approach calculated with sequentially designated waypoints in the absolute time. In this paper, we improved the method by utilizing the relative time expression for extension of the trajectory segment. Simulations with proposed method were done and the results were shown in this paper.

Rather than the difficulties of highly non-linear and non-Gaussian observation process and the state distribution in single target tracking, the presence of a large, varying number of targets and their interactions place more challenge on visual tracking. To overcome these difficulties, we formulate multiple targets tracking problem in a dynamic Markov network which consists of three coupled Markov random fields that model the following: a field for joint state of multi-target, one binary process for existence of individual target, and another binary process for occlusion of dual adjacent targets. By introducing two robust functions, we eliminate the two binary processes, and then apply a novel version of belief propagation called sequential stratified sampling belief propagation algorithm to obtain the maximum a posteriori (MAP) estimation in the dynamic Markov network. By using stratified sampler, we incorporate bottom-up information provided by a learned detector (e.g. SVM classifier) and belief information for the messages updating. Other low-level visual cues (e.g. color and shape) can be easily incorporated in our multi-target tracking model to obtain better tracking results. Experimental results suggest that our method is comparable to the state-of-the-art multiple targets tracking methods in several test cases.

Compressive Sensing Imaging (CSI) is a new framework for image acquisition, which enables the simultaneous acquisition and compression of a scene. Since the characteristics of Compressive Sensing (CS) acquisition are very different from traditional image acquisition, the general image compression solution may not work well. In this paper, we propose an efficient lossy compression solution for CS acquisition of images by considering the distinctive features of the CSI. First, we design an adaptive compressive sensing acquisition method for images according to the sampling rate, which could achieve better CS reconstruction quality for the acquired image. Second, we develop a universal quantization for the obtained CS measurements from CS acquisition without knowing any a priori information about the captured image. Finally, we apply these two methods in the CSI system for efficient lossy compression of CS acquisition. Simulation results demonstrate that the proposed solution improves the rate-distortion performance by 0.4~2 dB comparing with current state-of-the-art, while maintaining a low computational complexity.

This paper describes about experimental verifications on the cricket's response against frequent interactions and how it changes in time development. In order to investigate behavior modification, we took the approach to utilize automated interaction systems with a robotic agent and observe the reactions of the male cricket. The robotic agent robes cricket's pheromone and is controlled by designing motion patterns based on visual motion tracking of the both agents. We designed an automated interaction behavior pattern for the robotic agent and conducted interaction experiments to the male cricket. We observed that cricket showed ordinary response firstly and after a while the trend of the cricket's behavior changed. Thus, the response was somewhat modified in time development. In this paper, we explained experimental conditions and also showed the results of frequent interaction experiments by using a robotic agent.

Background prior selection is the key step in current ranking-based saliency detection approaches. The existing related methods usually choose boundary regions of an image or the region with low initial saliency value in single image scale as the background. Then, the saliency map is obtained by ranking the inside similarity and correlation. However, these methods cannot handle situations in which the salient object lies on the image boundary (boundary-salient) multiple salient objects exist in a single image (multisalient). To this end, this paper proposes an adaptive background selection method by exploiting the background invariance in different image scales within distinct color spaces. Through embedding the selected background prior into multiple newly proposed ranking-based saliency methods, the superiority of the obtained background prior is strongly verified. Exhaustive experiments on four challenging datasets demonstrate that the proposed method outperforms the state-of-the-art methods in handling the boundary-salient and multisalient situations.

This paper describes a path planning method for mobile robots/vehicles be extending the path sequentially. Considering practical use and applications of the mobile robot/vehicles, additional via-points for future path are given sequentially from sensor processing and environment recognition process. Thus, the path should be calculated sequentially with such additional conditions during mission execution. In this paper, we described a key idea of our method to extend the path iteratively. Polynomial-based time curve is utilized as the basis of the method and it describes path extension segment for certain period. The coefficients of such path extension segment are renewed by additional via-points. Computer simulations with proposed method were conducted and reported for fundamental confirmations. The results shows sequential path extension can be done by our proposed method.

This paper proposes probability iterative closest point (ICP) method based on expectation maximization (EM) estimation for point set registration with noise. The classical ICP algorithm can deal with rigid registration between two point sets effectively, but always fails to register point sets with noise. In order to improve the registration precision, a Gaussian model is introduced into the rigid registration. In each iteration, the classical ICP algorithm includes two steps, building the corresponding relationship and computing the rigid transformation. Similar to the traditional ICP, at each step, firstly the corresponding relationship is set up. Secondly, the rigid transformation is solved by singular value decomposition (SVD) method, and then the Gaussian model is updated by the distance and variance between two point sets. The experimental results on part B of CE-Shape-1 database and real position dataset validate that the proposed algorithm is more accurate.

In this paper, we propose an accurate and real- time positioning method for intelligent road vehicles in urban environments. The proposed method uses a robust lane marking detection algorithm, as well as an efficient shape registration algorithm between the detected lane markings and a GPS based road shape prior, to improve the robustness and accuracy of global localization of a road vehicle. We exploit both the state-of-the-art technologies of visual localization based on lane marking detection and the wide availability of Global Positioning System (GPS) based localization. We show that by formulating the positioning problem in a relative sense, we can estimate the vehicle localization in real-time and bound its absolute error in centimeter-level by a cross validation scheme. The validation scheme integrates the vision based lane marking detection with the shape registration, and improves the performance of the overall localization system. The GPS localization can be refined by using lane marking detection when the GPS suffers from frequent satellite signal masking or blockage, while lane marking detection is validated and completing by the GPS based road shape prior when it does not work well in adverse weather conditions or with poor lane signature. We extensively evaluate the proposed method with a single forward-looking camera mounted on an autonomous vehicle which travels at 60km/h through several urban street scenes.

The segmentation of categorized objects addresses the problem of joint segmentation of a single category of object across a collection of images, where categorized objects are referred to objects in the same category. Most existing methods of segmentation of categorized objects made the assumption that all images in the given image collection contain the target object. In other words, the given image collection is noise free. Therefore, they may not work well when there are some noisy images which are not in the same category, such as those image collections gathered by a text query from modern image search engines. To overcome this limitation, we propose a method for automatic segmentation and recognition of categorized objects from noisy Web image collections. This is achieved by cotraining an automatic object segmentation algorithm that operates directly on a collection of images, and an object category recognition algorithm that identifies which images contain the target object. The object segmentation algorithm is trained on a subset of images from the given image collection which are recognized to contain the target object with high confidence, while training the object category recognition model is guided by the intermediate segmentation results obtained from the object segmentation algorithm. This way, our co-training algorithm automatically identifies the set of true positives in the noisy Web image collection, and simultaneously extracts the target objects from all the identified images. Extensive experiments validated the efficacy of our proposed approach on four datasets: 1) the Weizmann horse dataset, 2) the MSRC object category dataset, 3) the iCoseg dataset, and 4) a new 30-categories dataset including 15,634 Web images with both hand-annotated category labels and ground truth segmentation labels. It is shown that our method compares favorably with the state-of-the-art, and has the ability to deal with noisy image collections.

This paper proposed an on-line reference governor for a mobile robot with velocity feedback controller. In order to follow given time path by a mobile robot, it is a practical subject how to cope with the disturbances by the slippage and the restrictions on the control system. In this paper, we proposed a methodology to adjust iteratively the control reference based on time scaling approach for a mobile robot to follow the path. By introducing proposed on-line reference governor, control input saturation can be reduced in case of that the disturbances occur during traveling and path tracking error is also reduced. For verification of our proposed method, computer simulations utilizing a plant with stable feedback controller were conducted and the results were discussed in this paper.

This paper describes the underwater data collection utilizing stationary sensor node and movable sensor node. We consider applying the system to investigate the conditions of marine ecosystem around the coral reef at Okinawa, Japan. The framework and fundamental design of our proposed underwater information gathering system under development are introduced. It is also reported the development of the prototypes of the sensory node and movable sensor node. Basic examinations utilizing such prototype at underwater environment are also described. This paper also describes experimental results of image gathering from via wireless local area network communication. Our future trials based on the current developmental situation are also discussed.

This paper proposed a design of reference governor for a mobile platform. To follow given time path by a mobile platform, it is a practical subject how to cope with the disturbances and limitation on the control plant. Proposed method is to adjust iteratively the control reference based on time scaling approach for a mobile platform. By introducing proposed reference governor, path tracking error can be reduced in case of that the disturbances occur during traveling. For verification of our proposed method, computer simulations were conducted.

Video object co-segmentation refers to the problem of simultaneously segmenting a common category of objects from multiple videos. Most existing video co-segmentation methods assume that all frames from all videos contain the target objects. Unfortunately, this assumption is rarely true in practice, particularly for large video sets, and existing methods perform poorly when the assumption is violated. Hence, any practical video object co-segmentation algorithm needs to identify the relevant frames containing the target object from all videos, and then co-segment the object only from these relevant frames. We present a spatiotemporal energy minimization formulation for simultaneous video object discovery and co-segmentation across multiple videos. Our formulation incorporates a spatiotemporal auto-context model, which is combined with appearance modeling for superpixel labeling. The superpixel-level labels are propagated to the frame level through a multiple instance boosting algorithm with spatial reasoning (Spatial-MILBoosting), based on which frames containing the video object are identified. Our method only needs to be bootstrapped with the frame-level labels for a few video frames (e.g., usually 1 to 3) to indicate if they contain the target objects or not. Experiments on three datasets validate the efficacy of our proposed method, which compares favorably with the state-of-the-art.

This paper describes an automated interaction for triggering the cricket's pheromone behavior by controlling a micro mobile robot based on pose estimation. In our developing systems a micro mobile robot : Ecobe mounted a cricket s head is introduced as a physical agent to interact with the cricket. Both of location of the cricket and mobile robot is estimated by on-line particle filter based color tracking and the pose estimation is also done by utilizing visual processing. By utilizing extracted information and novel designed robot's behavior algorithm, automated experimental trials in interaction between a cricket and a micro mobile robot has been done. Experimental results by utilizing our prototype system are also reported in this paper. Finally, pose estimation was also applied to the cricket and we show the result of fundamental trial of that.

Compressive sensing imaging (CSI) is a new framework for image coding, which enables acquiring and compressing a scene simultaneously. The CS encoder shifts the bulk of the system complexity to the decoder efficiently. Ideally, implementation of CSI provides lossless compression in image coding. In this paper, we consider the lossy compression of the CS measurements in CSI system. We design a universal quantizer for the CS measurements of any input image. The proposed method firstly establishes a universal probability model for the CS measurements in advance, without knowing any information of the input image. Then a fast quantizer is designed based on this established model. Simulation result demonstrates that the proposed method has nearly optimal rate-distortion (R~D) performance, meanwhile, maintains a very low computational complexity at the CS encoder.

Establishing correct correspondences between two images has a wide range of applications, such as 2D and 3D registration, structure from motion, and image retrieval. In this paper, we propose a new matching method based on spatial constraints. The proposed method has linear time complexity, and is efficient when applying it to image retrieval. The main assumption behind our method is that, the local geometric structure among a feature point and its neighbors, is not easily affected by both geometric and photometric transformations, and thus should be preserved in their corresponding images. We model this local geometric structure by linear coefficients that reconstruct the point from its neighbors. The method is flexible, as it can not only estimate the number of correct matches between two images efficiently, but also determine the correctness of each match accurately. Furthermore, it is simple and easy to be implemented. When applying the proposed method on re-ranking images in an image search engine, it outperforms the-state-of-the-art techniques.

Forensic detection of resampled digital images has become an important technology among many others to establish the integrity of digital visual content. This paper proposes a moment feature based method to detect resampled digital images. Rather than concentrating on the positions of characteristic resampling peaks, we utilize a moment feature to exploit the periodic interpolation characteristics in the frequency domain. Not only the positions of resampling peaks but also the amplitude distribution is taken into consideration. With the extracted moment feature, a trained SVM classifier is used to detect resampled digital images. Extensive experimental results show the validity and efficiency of the proposed method.

This paper presents a more robust iterative closest point (ICP) approach for 2D point set registration. An inequality constraint of the rotation angle is introduced into the least square registration model which is solved by an extended ICP algorithm. At each iterative step of the algorithm, a closed-form solution for the rotation is obtained according to the monotonicity of the model with respect to the rotation angle. The proposed approach extends the convergence domain of the ICP algorithm, and it can be used much more widely. A series of 2D point set experiments on part B of MPEG-7 CE-shape-1 dataset prove that the proposed method is much more robust than ICP without increasing the computational complexity.

Real-Time road detection is a demanding task in active safety and auto-driving of vehicle. Vision is the most popular sensing method for road detection, but it is easier to be influenced by illumination, shadows, shield, etc. To overcome those difficulties, the light detection and ranging (Lidar) sensor is a good choice for road detection. For either the urban environment or the rural areas, the important feature of the road is that the road surface could be approximately represented by some planes. Hence, the Lidar’s scanning plane and the road surface intersect at a set of line segments, and a line segment means a road plane. To extract the line segments from a scan, a least mean square problem is proposed, which is solved by a distance segment approach and an iterative line fitting approach. To eliminate the perception dead zone, some suitable historical data are adopted combining with the fresh data. In order to initial the road search range, a hypothesis is given that the area under the vehicle is road. The road detection is achieved for the scans from the close to the distance and in the meanwhile the search range of the next scan is updated. A lot of experiment results demonstrate the robustness and efficiency of the proposed approach for real-time auto-driving of the intelligent vehicles.

Salient object perception is the process of sensing the salient information from the spatio-temporal visual scenes, which is a rapid pre-attention mechanism for the target location in a visual smart sensor. In recent decades, many successful models of visual saliency perception have been proposed to simulate the pre-attention behavior. Since most of the methods usually need some ad hoc parameters or high-cost preprocessing, they are difficult to rapidly detect salient object or be implemented by computing parallelism in a smart sensor. In this paper, we propose a novel spatio-temporal saliency perception method based on spatio-temporal hypercomplex spectral contrast (HSC). Firstly, the proposed HSC algorithm represent the features in the HSV (hue, saturation and value) color space and features of motion by a hypercomplex number. Secondly, the spatio-temporal salient objects are efficiently detected by hypercomplex Fourier spectral contrast in parallel. Finally, our saliency perception model also incorporates with the non-uniform sampling, which is a common phenomenon of human vision that directs visual attention to the logarithmic center of the image/video in natural scenes. The experimental results on the public saliency perception datasets demonstrate the effectiveness of the proposed approach compared to eleven state-of-the-art approaches. In addition, we extend the proposed model to moving object extraction in dynamic scenes, and the proposed algorithm is superior to the traditional algorithms.

Localizationis of vital importance for an unmanned vehicle to drive on the road. Most of the existing algorithms are based on laser range finders, inertial equipment, artificial landmarks, distributing sensors or global positioning system(GPS) information. Currently, the problem of localization with vision information is most concerned. However, vision-based localization techniquesare still unavailable for practical applications. In this paper, we present a vision-based sequential probability localization method. This method uses the surface information of the roadside to locate the vehicle, especially in the situation where GPS information is unavailable. It is composed of two step, first, in a recording stage, we construct a ground truthmap with the appearance of the roadside environment. Then in an on-line stage, we use a sequential matching approach to localize the vehicle. In the experiment, we use two independent cameras to observe the environment, one is left-orientated and the other is right. SIFT features and Daisy features are used to represent for the visual appearance of the environment. The experiment results show that the proposed method could locate the vehicle in a complicated, large environment with high reliability.

This paper describes about a prototype of active interaction experiment system between a cricket and an operated micro mobile robot and measured/collected data in real-time by using the system. The behavior selection of the cricket (Gyllus bimaculatus) is influenced by the experience or the context in living environment. Therefore, we are trying to investigate neuronal mechanisms underlying micro brain of the cricket. For gathering behavioral data, we are developing a control/measurement system for realizing active interaction experiment. The prototype is composed of a micro mobile robot as a physical interaction agent, a camera and a microphone and a computer commands to the micro mobile robot and record the data of video sequence, motion tracking and the audio. Experimental trial using the prototype was done and reported.

This Letter proposes a novel saliency detection method based on biological plausibility of a hypercomplex Fourier spectrum contrast algorithm. The proposed algorithm takes into consideration not only simulation of simple cortical cells in the receptive field of humans but also the texture-color feature global spectrum contrast of an image. First, we utilize log-Gabor filters to mimic simple cortical cells in the receptive field of humans. Two complex numbers of texture colors are acquired through feature maps in hue, saturation, and value color space by log-Gabor. Second, we build the hypercomplex number using these representations of feature maps. Finally, the salient object is detected by spectrum contrast in the hypercomplex Fourier domain. Experimental results show that the proposed algorithm outperforms the state-of-the-art methods.

Saliency detection is an useful technique for image semantic analysis such as auto image segmentation, image retargeting, advertising design and image compression. Inspired by two existing saliency detection algorithms, named spectral residual (SR) and phase spectrum of quaternion Fourier transform (PQFT), we propose a new bottom-up saliency detection method which is featured with the introduction of hypercomplex spectral contrast (HSC) in saliency detection. The proposed HSC algorithm introduces the HSV color image vector space in hypercomplex number, and is better comprehensive to consider amplitude spectral contrast into saliency model as well as phase spectral contrast. Meanwhile, we also incorporate the human vision nonuniform sampling into our model, which is a common phenomenon that directs visual attention to the logarithmic center of image in natural scenes. Experimental results on two public saliency detection datasets show that our approach performs better than four state-of-the art approaches remarkably.

In this paper, a new blind and spatial-temporal feature based video watermarking algorithm for H.264/AVC is proposed. The novelty of the scheme lies in calculating an average frame which can represent the spatial-temporal characteristic of video, and creating a compound watermark by uniting the copyright information with average frame. This makes the watermark secure and robust. The other novel feature of the scheme is decomposing compound watermark into 8 bit-planes, and then embedding these bit-planes into different video segment. In watermarking detection, if the most significant bit-plane can be extracted, the copyright information can be recovered. As a result, it increases the watermark capacity while resists the frame dropping attacks. Furthermore, the compressed video bit rate is only increased 0.7% on average. Experimental results demonstrate the robustness and effectiveness of the proposed scheme.

We present a method for automatically extracting salient object from a single image, which is cast in an energy minimization framework. Unlike most previous methods that only leverage appearance cues, we employ an auto-context cue as a complementary data term. Benefitting from a generic saliency model for bootstrapping, the segmentation of the salient object and the learning of the auto-context model are iteratively performed without any user intervention. Upon convergence, we obtain not only a clear separation of the salient object, but also an auto-context classifier which can be used to recognize the same type of object in other images. Our experiments on four benchmarks demonstrated the efficacy of the added contextual cue. It is shown that our method compares favorably with the state-of-the-art, some of which even embraced user interactions.

The iterative closest point (ICP) algorithm is an accurate approach for the registration between two point sets on the same scale. However, it can not handle the case with different scales. This paper proposes a fast and robust ICP algorithm for isotropic scaling point sets registration (FRISICP). In order to accurately and directly estimate the scale factor without any constraints, we introduce a bidirection distance measurement method into the least square (LS) problem. Then to keep computational efficiency when the number of points in the set increasing, we further introduce a sparse-to-dense hierarchical model in ICP algorithm to speed up the isotropic scaling point set matching process. Experimental results demonstrate that the proposed FRISICP method outperforms other algorithms on both 2D and 3D point sets.

In this paper, we present a method to extract moving objects in monocular image sequences. The proposed method is based on graph cuts defined on a spatio-temporal region adjacency graph (RAG). First, we initially over-segment each frame in the video, and take the over-segmented regions as the vertices in the 3D spatio-temporal graph. Second, multiple cues are fused together to extract objects accurately. Finally, accurate foreground/background segmentation are efficiently achieved by binary graph cut. The experimental results showed that the proposed method improved the performance of segmentation with respect to the popular methods.

This paper proposes an efficient method to estimate the point spread function (PSF) of a blurred image using image gradients spatial correlation. A patch-based image degradation model is proposed for estimating the sample covariance matrix of the gradient domain natural image. Based on the fact that the gradients of clean natural images are approximately uncorrelated to each other, we estimated the autocorrelation function of the PSF from the covariance matrix of gradient domain blurred image using the proposed patch-based image degradation model. The PSF is computed using a phase retrieval technique to remove the ambiguity introduced by the absence of the phase. Experimental results show that the proposed method significantly reduces the computational burden in PSF estimation, compared with existing methods, while giving comparable blurring kernel.

Exposing digital forgeries by detecting local correlation patterns of images has become an important kind of approach among many others to establish the integrity of digital visual content. However, this kind of method is sensitive to JPEG compression, since compression attenuates the characteristics of local correlation pattern introduced by color filter array (CFA) interpolation. Rather than concentrating on the differences between image textures, we calculate the posterior probability map of CFA interpolation with compression related Gaussian model. Thus our approach will automatically adapt to compression. Experimental results on 1000 tampered images show validity and efficiency of the proposed method.

Human visual system function is non-uniform to visual perception. Normally, people tend to focus on one or more specific moving objects while watching video. In addition, video transmission is flawed due to heterogeneous networks and limited bandwidth. Therefore, how to transmit the content of interests more efficiently becomes a growing concerning topic. This paper proposes a method of multi-region in the scalable video coding. In the proposed method, the multi-region firstly get through the particle tracking filters, and then it is implemented motion estimation to each region and finally allocated bit-rate under the 3D-wavelet scalable video coding model. Experimental results show that the performance of our approach is effective, in particularly the situations of lower bit rate. The approach is particularly suitable for mobile video communication, and video surveillance on intelligent vehicle.

This paper describes a cluster-based transcoder that could transcode MPEG-2 format video files to MPEG-4 and H.264 format video files with faster transcoding speed than traditional single PC transcoder1. A seamless method of video division and a novel multi-resolution load balancing algorithm for distributed transcoding are proposed, and meanwhile, a fast algorithm for embedding watermarking is presented. The experimental results show that the proposed cluster-based digital video transcoder provides a good trade-off in terms of speed and quality compared with previous popular methods.

This paper proposes a new covariance matching based technique for blurred image PSF (point spread function) estimation. A patch based image degradation model is proposed for the covariance matching estimation framework. A robust covariance metric which is based on Riemannian manifold is adapted to measure the distance between covariance matrices. The optimal PSF is computed by minimizing the dissimilarity between the covariance matrix of the blurred image patches and the theoretical covariance matrix which is based on autoregressive image model. Experimental results show that the proposed method produces promising PSF estimation results.

We present a hierarchical and compositional model based on an And-or graph for joint detecting and tracking of multiple targets in video. In the graph, an And-node for the joint state of all targets is decomposed into multiple Or-nodes. Each Or-node represents an individual target's state that includes position, appearance, and scale of the target. Leaf nodes are trained detectors. Measurements that supplied by the predictions of the tracker and leaf nodes are shared among Or-nodes.There are two kinds of production rules respectively designed for the problems of varying number and occlusions. One is association relations that distributes measurements to targets, and the other is semantic relations that represent occlusion between targets. The inference algorithm for the graph consists of three processing channels: (1) a bottom-up channel, which provides informative measurements by using learned detectors; (2) a top-down channel, which estimates the individual target state with joint probabilistic data association; (3) a context sensitive reasoning channel, which finalizes the estimation of the joint state with belief propagation. Additionally, an interaction mechanism between detection and tracking is implemented by a hybrid measurement process. The algorithm is validated widely by tracking peoples in several complex scenarios. Empirical results show that our tracker can reliably track multi-target without any prior knowledge about the number of targets and the targets may appear or disappear anywhere in the image frame and at any time in all these test videos.

An arbitrary shape region of interest (ROI) coding is presented for scalable wavelet video codec in this paper. The padding of macroblock and polygon matching are employed to estimate the motion of the ROI. The motion vectors derived are set as the motion trajectory of the samples to generate one-dimensional temporal signal, filtered to reduce the temporal redundancy by using motion compensated temporal filtering for arbitrary shape ROI. The Reconstructed quality of the ROI coding can be significantly improved at low bit rate, compared to non-ROI coding. The efficiency of the MCTF based on arbitrary ROI is compared with that of the video object coding in MPEG-4. The ability of the MCTF to reduce the temporal redundancy is better than or comparable to that MPEG-4 to some extent.

Joint network-source video coding (JNSC) is targeted to achieve the optimum delivery of a video source to a number of destinations over network with capacity constraints. In this paper, a practical scalable multiple description coding is proposed for JNSC, based on Lagrangian rate allocation and scalable video coding. After the spatiotemporal wavelet transformation of input video sequence and the bit plane coding and context-based adaptive binary arithmetic coding, jointing network-source coding is performed on the coding passes of the code blocks using Lagrangian rate allocation. The principle relationship of the rate-distortion slop ratio with receiving probability in network links is derived under the link capacity constraints. In this way, scalable multiple descriptions can be generated to optimize the delivery to be robust and adaptive to the dynamics of heterogeneous networks. The performance of the proposed scalable multiple description coding is explored in the Peer-to-Peer streaming network.

Motion analysis involves the interpretation of image data over time. It is crucial for a range of vision tasks such as obstacle detection, depth estimation, video analysis, scene interpretation, video compression, etc. Motion analysis is difficult because it requires modeling of the complicated relationships between the observed image data and the motion of objects and motion patterns in the visual scene.

In this chapter, we discuss, in depth, visual information processing and a new artificial intelligent (AI) system that is based upon cognitive mechanisms. The relationship between a general model of intelligent systems and cognitive mechanisms is described, and in particular we explore visual information processing with selective attention. We also discuss a methodology for studying the new AI system and propose some important basic research issues that have emerged in the intersecting fields of cognitive science and information science. To this end, a new scheme for associative memory and a new architecture for an AI system with attractors of chaos are addressed.

This chapter presents the fundamentals of statistical pattern recognition and statistical learning. First, we present the general framework of a statistical pattern recognition system and discuss pattern representation and classification, two important components of such a system. Second, we introduce the concept of statistical learning and examine the three main approaches to statistical learning: supervised learning, semi-supervised learning, and unsupervised learning.