This paper proposes a unified observer–identifier–learning framework (OILF) for predefined-time optimal tracking control with prescribed performance for robotic systems subject to unmeasurable states and uncertain dynamics. Most existing optimal control approaches rely on full-state information or accurate dynamic models, which are often unavailable in practice. To overcome this issue, a predefined-time dynamic regression extension and mixing (PTDREM) method is proposed to realize the co-design of state observer and parameter identifier, enabling synchronous predefined-time estimation of unknown states and dynamic parameters. Subsequently, to achieve optimal tracking control for robotic systems, a prescribed-performance-based critic–actor (PPCA) structure is developed via reinforcement learning (RL), in which all hierarchical tracking errors are driven into prescribed neighborhoods of the origin within a predefined time. In contrast to most existing works that solely ensure uniform ultimate boundedness (UUB) of the closed-loop system, the proposed scheme enables the upper bounds of the convergence time of the state observer, system identifier, and optimal controller to be preset through independent parameter design, thereby establishing global predefined-time stability (G-PTS) for the overall closed-loop system. Numerical simulations on a two-degree-of-freedom (DOF) robotic manipulator verify the effectiveness of the proposed OILF.

In this article, a novel general clustered network framework with hybrid communication is constructed for multiple uncertain Euler-Lagrange (EL) systems. The objective is to ensure that the systems under consideration achieve containment tracking in the predefined path at the prescribed time. First, in the light of the hierarchical control design, the output tracking problem is decomposed into the desired signals tracking and the stability of the nonlinear uncertain coupled systems. Second, the wide-area network is described by a combination of a directed graph and an intermittent control scheme, then all agents are divided into different subnetworks in demand or scenarios. Based on this, a distributed prescribed-time hybrid observer under a time-varying scaling function and a novel containment error method is designed to achieve the containment tracking. In addition, an adaptive distributed prescribed-time hybrid control strategy is proposed for the uncertainty estimation. Then, the prescribed-time stability of uncertain EL systems is analyzed and guaranteed using the general Lyapunov theory and intermittent control method. Finally, the proposed hybrid control strategy is verified by the simulation results of multiple flexible manipulator systems.

Motor imagery-based brain–computer interfaces (MI-BCIs) hold significant promise for rehabilitation training in individuals with neurological impairments such as stroke and spinal cord injury (SCI). Achieving precise and robust lower limb movement prediction for each patient is crucial. However, the variability in MI response frequencies and brain activation patterns among subjects presents a great challenge to the generalizability of MI-BCIs. This paper proposes a Tuned Heuristic Fusion Graph Convolutional Network (THFGCN) for limb movement prediction in rehabilitation scenarios. THFGCN innovatively designs a learnable EEG frequency band tuned module and a heuristic space topology module. These two modules allow for the intricate extraction of both frequency and spatial topological features, utilizing graph adjacency matrices that encapsulate channel correlations and spatial relationships, hence fostering individualized analysis and enhanced generalizability across subjects. Furthermore, a spatio-temporal convolution module paired with a feature map attention mechanism is proposed to extract the critical spatio-temporal features of electroencephalogram (EEG) data. Validation experiments on the PhysioNet and LLM-BCImotion datasets against six mainstream methods demonstrate that THFGCN outperforms state-of-theart methods, achieving 88.41% and 82.82% accuracy in the within-subject case, and 65.93% and 60.56% accuracy in the cross-subject case, respectively. Detailed frequency band weight and T-distributed Stochastic Neighbor Embedding visualization validate the effectiveness of proposed modules. Furthermore, feature interpretability analysis proves the extracted features’ profound MI task relevance, underlining THFGCN’s exceptional interpretability.

This study investigates the resilient output containment problem in heterogeneous multiagent systems that facing intermittent communication and sensor attacks, specifically focusing deception attacks within wide-area networks (WANs). A novel WAN framework is introduced to effectively analyze the intricate nature of the network. The proposed framework facilitates the examination of connected agents across multiple scenarios to achieve the desired containment objective. The framework involves three key steps. First, a distributed hybrid control strategy is developed, utilizing an internal model to handle intermittent communication between clusters and continuous communication within clusters. Second, considering the unknown states of compromised heterogeneous agents, a Luenberger observer is devised for each agent, employing adaptive observers to estimate the state and external attack information. Subsequently, a distributed hybrid controller is proposed to achieve global output containment, ensuring uniform ultimate boundedness. Additionally, a sufficient condition for exponential stability is derived to tackle the intermittent control problem. This criterion employs the characterization of average intermittent intervals. The effectiveness of the proposed adaptive hybrid control strategy is demonstrated through simulation examples, showcasing its ability to address the challenges posed by intermittent communication and deception attacks.

In this paper, recently studied rigid body rotation control problems are summarized as a constrained dynamics on SO(3) from the point of view of a nonlinear multi input multi output system. Our specific interests are in the rotation of human head and eye, where the control objective is to direct the pointing direction towards a stationary or mobile point target in space. Eye and head rotation control problems were studied in the mid-nineteenth century by Listing, Donders, Helmholtz and others. In these studies, it was observed by Donders that, for rotations away from the primary direction, the rotation vectors are restricted to lie on a surface, called the Donders' surface. Additionally, for eye movements under head-fixed condition, it was observed by Listing that the Donders' surface is actually a plane, called the Listing's plane. The input signals to the eye and head are provided by a triplet of external torques generated by muscles. Three output signals from the eye and head are chosen as follows. Two of the signals are coordinates of the frontal pointing direction. The third signal measures deviation of the state vector from the Donders' surface (respectively Listing's plane). Thus we have a 3 × 3 square system and the claim is that the proposed square system is locally feedback linearizable on a suitable neighborhood N of the state space.

In this paper, an event-triggered optimal tracking control of discrete-time multi-agent systems is addressed by using reinforcement learning. In contrast to traditional reinforcement learning-based methods for optimal coordination and control of multi-agent systems with a time-triggered control mechanism, an event-triggered mechanism is proposed to update the controller only when the designed events are triggered, which reduces the computational burden and transmission load. The stability analysis of the closed-loop multi-agent systems with event-triggered controller is described. Further, to implement the proposed scheme, an actor-critic neural network learning structure is proposed to approximate performance indices and to on-line learn the event-triggered optimal control. During the training process, event-triggered weight tuning law has been designed, wherein the weight parameters of the actor neural networks are adjusted only during triggering instances compared with traditional methods with fixed updating periods. Further, a convergence analysis of the actor-critic neural network is provided via Lyapunov method. Finally, two simulation examples show the effectiveness and performance of the obtained event-triggered reinforcement learning controller.

In this article, a novel reinforcement learning (RL) method is developed to solve the optimal tracking control problem of unknown nonlinear multiagent systems (MASs). Different from the representative RL-based optimal control algorithms, an internal reinforce Q-learning (IrQ-L) method is proposed, in which an internal reinforce reward (IRR) function is introduced for each agent to improve its capability of receiving more long-term information from the local environment. In the IrQL designs, a Q-function is defined on the basis of IRR function and an iterative IrQL algorithm is developed to learn optimally distributed control scheme, followed by the rigorous convergence and stability analysis. Furthermore, a distributed online learning framework, namely, reinforce-critic-actor neural networks, is established in the implementation of the proposed approach, which is aimed at estimating the IRR function, the Q-function, and the optimal control scheme, respectively. The implemented procedure is designed in a data-driven way without needing knowledge of the system dynamics. Finally, simulations and comparison results with the classical method are given to demonstrate the effectiveness of the proposed tracking control method.

In this study, a finite-time velocity-free rendezvous control method is considered for multiple autonomous underwater vehicle (AUV) systems with intermittent undirected communication. First, we develop a distributed finite-time observer for each AUV to estimate its own state information. Second, we design a rendezvous control algorithm that utilizes the estimated state information intermittently through a communication network in the absence of velocity measurement. A homogeneous method is used to prove that all AUVs in the group can achieve rendezvous in finite time for a network with intermittent communication, even without velocity measurements. The proposed method is shown to reduce the communication load of the system. More importantly, the control algorithm achieves the control goal of the system and is proven to be viable for many practical applications of multiple AUV systems from both economic and security perspectives. Finally, the effectiveness of the proposed control protocol is demonstrated via numerical simulations.

This article investigates an output antisynchronization problem of multiagent systems by using an input-output data-based reinforcement learning approach. Till now, most of the existing results on antisynchronization problems required full-state information and exact system dynamics in the controller design, which is always invalid in practical scenarios. To address this issue, a new system representation is constructed by using just the available input/output data from the multiagent system. Then, a novel value iteration algorithm is proposed to compute the optimal control laws for the agents; moreover, a convergence analysis is presented for the proposed algorithm. In the implementation of the data-based controllers, an actor-critic network structure is established to learn the optimal control laws without the requirement of information of the agent dynamics. An incremental weight updating rule is proposed to improve the learning performance. Finally, simulation results are presented to demonstrate the effectiveness of the proposed antisynchronization control strategy.

Lower limb exoskeleton (LLE) has received considerable interests in strength augmentation, rehabilitation and walking assistance scenarios. For walking assistance, the LLE is expected to have the capability of controlling the affected leg to track the unaffected leg’s motion naturally. An important issue in this scenario is that the exoskeleton system needs to deal with unpredictable disturbance from the patient, which requires the controller of exoskeleton system to have the ability to adapt to different wearers. This paper proposes a novel Data-Driven Reinforcement Learning (DDRL) control strategy to adapt different hemiplegic patients with unpredictable disturbances. In the proposed DDRL strategy, the interaction between two lower limbs of LLE and the legs of hemiplegic patient are modeled in the context of leader-follower framework. The walking assistance control problem is transformed into a optimal control problem. Then, a policy iteration (PI) algorithm is introduced to learn optimal controller. To achieve online adaptation control for different patients, based on PI algorithm, an Actor-Critic Neural Network (ACNN) technology of the reinforcement learning (RL) is employed in the proposed DDRL. We conduct experiments both on a simulation environment and a real LLE system. Experimental results demonstrate that the proposed control strategy has strong robustness against disturbances and adaptability to different pilots.

This paper discussed the recent advances in Deep Sequencing Data Analysis for systems biology research. Deep sequencing technologies have been primarily applied to genomic sequencing but have recently been applied for transcriptomic profiling or mapping histone modifications. Deep Sequencing technology shows clear advantages over existing profiling technologies in terms of amount of sequence coverage, revealing new transcriptomic insights, measurement of expression of different transcript isoforms and accuracy of defining transcription level. However, being a relatively newer method for transcriptomic profiling, standardized approaches for analysis of deep sequencing expression data are still being developed. The analysis and application of deep sequencing data presents enormous challenges in the areas of machine learning, signal processing, systems theory and statistics. The emphasis of the special issue is on the latest computational challenges and finding rigorous and novel engineering approaches to tackle structural and functional systems biology problems using deep sequencing technologies

A major challenge to solve problems in control of Boolean networks is that the computational cost increases exponentially when the number of nodes in the network increases. We consider the problem of controllability and stabilizability of Boolean control networks, address the increasing cost problem by partitioning the network graph into several subnetworks, and analyze the subnetworks separately. Easily verifiable necessary conditions for controllability and stabilizability are proposed for a general aggregation structure. For acyclic aggregation, we develop a sufficient condition for stabilizability. It dramatically reduces the computational complexity if the number of nodes in each block of the acyclic aggregation is small enough compared with the number of nodes in the entire Boolean network.

Behavior of living organisms is strongly modulated by the day and night cycle giving rise to a cyclic pattern of activities. Such a pattern helps the organisms to coordinate their activities and maintain a balance between what could be performed during the "day" and what could be relegated to the "night." This cyclic pattern, called the "Circadian Rhythm," is a biological phenomenon observed in a large number of organisms. In this paper, our goal is to analyze transcriptome data from Cyanothece for the purpose of discovering genes whose expressions are rhythmic. We cluster these genes into groups that are close in terms of their phases and show that genes from a specific metabolic functional category are tightly clustered, indicating perhaps a "preferred time of the day/night" when the organism performs this function. The proposed analysis is applied to two sets of microarray experiments performed under varying incident light patterns. Subsequently, we propose a model with a network of three phase oscillators together with a central master clock and use it to approximate a set of "circadian-controlled genes" that can be approximated closely.

Life on earth is strongly affected by the day-night cycle, also known as the diurnal cycle. Due to its importance in survival, many organisms have developed an internal time keeping mechanism that goes by the name of circadian rhythm. Light plays a vital role for photosynthetic cyanobacteria and changes in the light pattern result in adaptive changes in the underlying biological processes at cellular level. Processes under circadian control are able to maintain their rhythm even under changes in the diurnal cycle, and it is important to isolate these processes from those whose rhythms are strongly affected by light. As the only known prokaryotic organism to have a robust circadian clock mechanism, cyanobacteria provide us with a unique opportunity to unveil the complex changes, especially changes in the process rhythms resulting from perturbation of the diurnal cycle. In this paper, we have identified the circadian controlled genes from those that are strongly influenced by light using a pair of genome wide study utilizing microarrays. A transcription model with an associated regulatory network is proposed.

Behavior of living organisms is strongly modulated by light especially by the day and night cycle giving rise to a cyclic pattern of activities. Such a pattern helps the organism to coordinate their activities and maintain a balance between what could be performed during the ‘day’ and what could be relegated to ‘night’. This cyclic pattern, called the ‘Circadian Rhythm’, is a biological phenomenon observed in a large number of organisms ranging from unicellular bacteria to human beings and is present in data collected at various levels viz. transcriptome, proteome etc. In this paper, our goal is to analyze transcriptome data from Cyanothece, a photosynthetic cyanobacteria, for the purpose of discovering genes whose expressions are rhythmic, especially those for which these rhythms have a 24 hours cycle. Subsequently we propose a model with a network of three phase oscillators for each one of the twenty four hours cycle. Each of the three phase oscillators is chosen to maintain a phase difference of 120 degrees between each other. All the oscillators are connected to an internal clock that is designed to maintain a phase activity close to a master clock derived using KaiC proteins. In Cyanobacteria it is believed that the KaiC proteins provide the internal rhythm. The model parameters, viz. connection strengths between the master clock and peripheral oscillators and the parameters computing the linear combinations of the oscillator phase variables, are optimized to provide a close match to the observed gene expressions even when the frequency of the internal clock and the natural frequencies of the oscillators vary within a certain range. As a final step, the oscillator network model has been used to isolate genes, and hence the associated subprocesses, whose expression cycles are robust with respect to variations in the oscillator frequencies.

Animals routinely rely on their eyes to localize fixed and moving targets. Such a localization process might include prediction of future target location, recalling a sequence of previously visited places or, for the motor control circuit, actuating a successful movement. Typically, target localization is carried out by fusing images from two eyes, in the case of binocular vision, wherein the challenge is to have the images calibrated before fusion. In the field of machine vision, a typical problem of interest is to localize the position and orientation of a network of mobile cameras (sensor network) that are distributed in space and are simultaneously tracking a target. Inspired by the animal visual circuit, we study the problem of binocular image fusion for the purpose of localizing an unknown target in space. Guided by the dynamics of “eye rotation,” we introduce control strategies that could be used to build machines with multiple sensors. In particular, we address the problem of how a group of visual sensors can be optimally controlled in a formation. We also address how images from multiple sensors are encoded using a set of basis functions, choosing a “larger than minimum” number of basis functions so that the resulting code that represents the image is sparse. We address the problem of how a sparsely encoded visual data stream is internally represented by a pattern of neural activity. In addition to the control mechanism, the synaptic interaction between cells is also subjected to “adaptation” that enables the activity waves to respond with greater sensitivity to visual input. We study how the rat hippocampal place cells are used to form a cognitive map of the environment so that the animal's location can be determined from its place cell activity. Finally, we study the problem of “decoding” location of moving targets from the neural activity wave in the cortex.

We study a new scheme for map building and describe localization techniques for a mobile robot equipped with a 2D laser rangefinder. We propose to use line segments as the basic element for the purpose of localization and to build the map. Line segments do provide considerable geometric information about the scene that can also be used for accurate and fast localization. We introduce a new closed line segment (CLS) map which consists of only line segments and defines a closed and connected region. Virtual line segments are drawn, for the spots that do not adequately describe a line segment on the range data. These are further explored via navigation and we argue that the CLS map provides an efficient mobile robot exploration scheme. All these techniques have been implemented on our Nomad XR4000 mobile robot and results are described in this paper.

A 2 1/2 D geometric feature based map is introduced. It is a closed map that integrates laser, sonar range data and bumper tactile data into a multi-layer map under a unified geometric description. We use line segments, circular arcs and point clusters as the basic geometric features. An Extended Kalman Filter (EKF) based on geometric features provides localization and the associated uncertainty. Virtual Line Segment Pursuit strategy is used for systematic exploration. Sensor placement planning is introduced for this Simultaneous Localization and Mapping (SLAM) problem and the motion is planned to improve the localization accuracy and target observability. Experimental results are shown to verify the effectiveness of the proposed scheme.

This paper analyzes the problem of shape modeling using the principle of active geometric deformable models. While the basic modeling technique already exists in the literature, we highlight many of its drawbacks and discuss their source and steps to overcome them. We propose a new stopping criterion to address the stopping problem. We also propose to apply a level set algorithm to implement the active geometric deformable models, thereby handling topology changes automatically. To alleviate the numerical problems associated with the implementation of the level set algorithm, we propose a new adaptive multigrid narrow band algorithm. All the proposed new changes have been illustrated with experiments with synthetic images and medical images.

This paper addresses the problem of motion estimation and selective reconstruction of objects undergoing rotational motion composed with translational motion. The goal is to derive the motion parameters belonging to the multiple moving objects, i.e. the angular velocities and the translational velocities and identify their locations at each time instance by selective reconstruction. These parameters and locations can be used for various purpose such as trajectory tracking, focus/shift attention of robot, etc. The innovative algorithm we have developed is based on angular velocity and translational velocity tuned 2D+T filters. One of the important facts about the algorithm is that it is effective for both spinning motion and orbiting motion, and thus unifies the treatment of the two kinds of rotational motion. Also by tuning of the filters, we can derive the translational motion parameters and the rotational motion parameters separately, which has the advantage of making motion estimation faster and more robust compared to estimating all of them simultaneously. The algorithm is simulated using synthesized image sequences corrupted by noise and is shown to be accurate and robust against noise and occlusion.

This paper addresses the problem of estimating, analyzing and tracking objects moving with spatio-temporal rotational motion (spin or orbit). It is assumed that the digital signals of interest are acquired from a camera and structured as digital image sequences. The trajectories in the signal are two-dimensional spatial projections in time of motion taking place in a three-dimensional space. The purpose of this work is to focus on the rotational motion, i.e. estimate the angular velocity. In natural scenes, rotational motion usually composes with translational or accelerated motion on a trajectory. This paper shows that trajectory parameters and rotational motion can be efficiently estimated and tracked either simultaneously or separately. The final goal of this work is to provide selective reconstructions of moving objects of interest. This paper constructs new continuous wavelet transforms that can be tuned to both translational and rotational motion. The parameters of analysis that are taken into account in these rotational wavelet transforms are space and time position, velocity, spatial scale, angular orientation and angular velocity. The continuous wavelet functions are finally discretized for signal processing. The link between rotational motion, symmetry and critical sampling is also presented. Applications are presented with tracking and estimation.

This paper addresses the problem of estimating and analyzing accelerated motion in spatio-temporal discrete signals. It is assumed that the digital signals of interest are acquired from imaging sensors and structured as digital image sequences. The motion trajectories in the signal are two-dimensional spatial projections in time of three-dimensional motions. Consequently, they contain all the orders of acceleration. The purpose of this work is to estimate the trajectory and the motion parameters of selected moving objects in the scene. The final goal is to provide selective reconstructions of accelerated objects of interest. This paper presents the construction of new continuous wavelet transforms that can be tuned to any order of accelerations, we demonstrate their existence and provide the related admissibility conditions. The parameters for analysis that are taken into account in these accelerated wavelet transforms are spatial and temporal translations, velocity, acceleration (second or nth order), spatial scale and spatial rotation. The continuous wavelet functions are finally discretized for signal processing.

The purpose of this paper is to develop a motion based segmentation for digital image sequences that is based on the continuous wavelet transform. The continuous wavelet transform allows estimating the motion parameters on all the moving discontinuities, edges and boundaries in the image sequence. This technique provides all the information of motion parameter estimates and edge locations at once without going back and forth refining the segmentation and the motion parameter estimation. Also, this is achieved without involving any point/block corresponding techniques in our algorithm. The edges and the motion parameter estimates are calculated locally on small windows or pixels in the image planes by maximizing the square of the modulus of the wavelet transform. A clustering procedure allows separating all the detected edges into clusters of homogeneous motion. Building a ridge skeleton on the reconstructed edges in each cluster provides the ultimate motion-based segments or partition. The algorithm was simulated using real traffic image sequences acquired by a mobile camera and proved to be accurate and robust.

This paper discusses the planning and control problems of a robot manipulator for a class of constrained motions. The task under consideration is to control a robot such that a tool grasped by the end-effector of the robot follows a path on an unknown surface with the aid of a single camera vision system. To accomplish the task, we propose a new planning and control strategy based on multisensor fusion. Three different sensors-joint encoders, a wrist force-torque sensor and a vision system with a single camera fixed above a workspace-are employed. We decouple control variables into two sub-spaces: one is for force control and the other for control of constrained motion. We also develop a new scheme by means of sensor fusion to handle the uncertainties in an uncalibrated workspace. The contact surface is assumed to be unknown and the precise position and orientation of the camera with respect to the robot is also unknown.

The real-time planning and control problems of a robot manipulator in an unstructured workspace are considered. Our goal is to control a robot to follow an unknown path. Difficulties arise when the robot is required to work in an unstructured workspace. Based on multisensor fusion, a novel strategy for integrating real-time planning and control is proposed to deal with the uncertainties in the environment. In the proposed scheme, a new hybrid position/force control strategy is utilized to maintain the contact between the robot and the unknown surface, while a new planner is designed to complete the path-following task without a priori knowledge of the path. In the paper, to achieve intelligent robot manipulation in an unstructured workspace, multi-sensor fusion is employed both for force-torque and visual sensors with complementary observed data compared to the traditional fusion schemes with redundant data.

The essence of our scheme is to extend our earlier (Ghosh et al., 1996) proposed multi-sensor based observation scheme to a situation where the location of the mobile platform is unknown. A new self-calibrated manipulation scheme is proposed to determine the position and orientation of the part with respect to the coordinate system attached to the base of the robot manipulator. The novelties of the proposed approach can be summarized as (i) multi-sensor fusion scheme based on complementary data for the purpose of part localization, (ii) part tracking and grasping control based on parallel tracking, and (iii) self-calibration of the location of the mobile platform using visual data and feature points on the end-effector. The principle advantages of the proposed scheme are described as follows. (i) It renders possible reconfiguring a manufacturing workcell without recalibrating the robot based coordinate frame. This significantly reduces the setup time of the workcell. (ii) It significantly reduces the requirement on the image processing speed. Experimental study of the proposed scheme has been carried out and the results are reported in this paper.

The main problem that we address in this paper is how a robot manipulator is able to track and grasp a part placed arbitrarily on a moving disc conveyor aided by a single CCD camera and fusing information from encoders placed on the conveyor and also from encoders on the robot manipulator. The important assumption that distinguishes our work from what has been previously reported in the literature is that the position and orientation of the camera and the base frame of the robot is a priori assumed to be unknown and is 'visually calibrated' during the operation of the manipulator. Moreover the part placed on the conveyor is assumed to be nonplanar, i.e. the feature points observed on the part is assumed to be located arbitrarily in R/sup 3/. The novelties of the proposed approach in this paper includes a (i) multisensor fusion scheme based on complementary data for the purpose of part localization, and (ii) self-calibration between the turntable and the robot manipulator using visual data and feature points on the end-effector. The principle advantages of the proposed scheme are the following. (i) It renders possible to reconfigure a manufacturing workcell without recalibrating the relation between the turntable and the robot. This significantly shortens the setup time of the workcell. (ii) It greatly weakens the requirement on the image processing speed.

In this paper we introduce a new approach to visually manipulate, with the aid of a robot manipulator, a part placed randomly on a rotating turntable. The highlight of our approach is that the camera and the robot end effector are both assumed to be uncalibrated. We only assume that the height of the robot end effector is known. Our approach utilizes virtual rotation of the camera via image processing not previously introduced in the literature. Finally, the tracking scheme is implemented by planning the error and gradually forcing it to zero while maintaining the torque controls within acceptable limits. This way we demonstrate a new visually guided analytical tracking scheme.

Discusses the problem of using visual and other sensors in the manipulation of a part by a robotic manipulator in a manufacturing workcell. The authors' emphasis is on the part localization problem involved. The authors introduce a new sensor-fusion approach which fuses sensory information from different sensors at various spatial and temporal scales. Relative spatial information obtained from processing of visual information is mapped to absolute taskspace of the robot through fusing of information from an encoder. Data obtained this way can be superimposed upon data obtained from displacement based vision algorithms at coarser time scales to improve overall reliability. Tracking plans reflecting sensor fusion are proposed. The localization of a part by spatial sensor fusion is experimentally demonstrated to be able to give required fast and accurate part localization.

A multi-sensor based planning and control scheme for robotic manufacturing is presented in this paper. The proposed approach fuses sensory information from various sensors at different temporal and spatial scales in an event-based planning and control scheme. By combining the measurement of an encoder sensor, relative spatial information obtained from processing of visual measurement is mapped to the absolute task-space of the robot, delayed data obtained from a displacement based vision algorithm that represent absolute part position measurement is brought to up to date. A four-step approach to planning and control of a robotic manipulator is discussed. An event-driven tracking and control scheme that is based on multi-sensor information is given. The approach is illustrated by considering a manufacturing workcell where the manipulator is commanded to pick up a part on a disc conveyor under the guidance of computer vision.

The multiresolution filtering technique is developed and applied to smooth image segmentation. The technique is based on approximations of a signal by some orthonormal wavelet basis at various resolutions. It is shown that the filter is close to being shift invariant and linear in phase according to an index on wavelets and also capable of suppressing local quadratic variations due to small bumps or noise by going to coarser approximations. An image segmentation scheme is proposed based on the technique. It has been tested on a variety of real images.>

This paper introduces identifiability problems that arise in linear dynamical systems with perspective observation function. Such a perspective problem finds its application in the field of Computer Vision specifically in the area of motion estimation of a rigid body with point, line or curve correspondence. The basic result of this paper is to study the correspondence problems in an unifying framework and it is shown that these problems arise as a special case of a more general Perspective System Problem. Problems in perspective system theory introduced in this paper are new.