Showing papers presented at "International Conference on Control, Automation, Robotics and Vision in 2012"

Face liveness detection by exploring multiple scenic clues

Abstract: Liveness detection is an indispensable guarantee for reliable face recognition, which has recently received enormous attention. In this paper we propose three scenic clues, which are non-rigid motion, face-background consistency and imaging banding effect, to conduct accurate and efficient face liveness detection. Non-rigid motion clue indicates the facial motions that a genuine face can exhibit such as blinking, and a low rank matrix decomposition based image alignment approach is designed to extract this non-rigid motion. Face-background consistency clue believes that the motion of face and background has high consistency for fake facial photos while low consistency for genuine faces, and this consistency can serve as an efficient liveness clue which is explored by GMM based motion detection method. Image banding effect reflects the imaging quality defects introduced in the fake face reproduction, which can be detected by wavelet decomposition. By fusing these three clues, we thoroughly explore sufficient clues for liveness detection. The proposed face liveness detection method achieves 100% accuracy on Idiap print-attack database and the best performance on self-collected face anti-spoofing database.

3D finger posture detection and gesture recognition on touch surfaces

Abstract: As the use of touch surfaces for user interfaces becomes more common, advances in the interpretations of touch input have lagged behind and are still limited to only the most basic of motions. Richer gesture-based human-computer interactions could serve to advance a wider acceptance of touch-based technology in a variety of fields. Using a 3D finger posture rather than just the 2D contact point in gesture definition opens the door to very rich, expressive, and intuitive gesture metaphors. In this paper, we present algorithms and methods for estimating the parameters of 3D finger postures on a touch surface, as well as a gesture recognition framework which uses an Artificial Neural Network to recognize 3D gestures on touchpads and touchscreens.

Electromyogram (EMG) feature reduction using Mutual Components Analysis for multifunction prosthetic fingers control

Abstract: Surface Electromyogram (EMG) signals are usually utilized as a control source for multifunction powered prostheses. A challenge that arises with the current demands of such prostheses is the ability to accurately control a large number of individual and combined fingers movements and to do so in a computationally efficient manner. As a response to such a challenge, we present a combined feature selection and projection algorithm, denoted as Mutual Components Analysis (MCA). The proposed MCA algorithm extends the well-known Principal Components Analysis (PCA) by pruning the noisy and redundant features before projecting the data. To implement the feature selection step, the mutual information concept is utilized to implement a new information gain evaluation function. The performance and significance of the proposed MCA is demonstrated on EMG datasets collected for the purpose of this research from eight subjects with eight electrodes attached on their forearm. Fifteen classes of fingers movements where considered in this paper with MCA achieving >95% accuracy on average across all subjects.

Detecting wandering behavior based on GPS traces for elders with dementia

Abstract: Wandering is among the most frequent, problematic, and dangerous behaviors for elders with dementia. Frequent wanderers likely suffer falls and fractures, which affect the safety and quality of their lives. In order to monitor outdoor wandering of elderly people with dementia, this paper proposes a real-time method for wandering detection based on individuals' GPS traces. By representing wandering traces as loops, the problem of wandering detection is transformed into detecting loops in elders' mobility trajectories. Specifically, the raw GPS data is first preprocessed to remove noisy and crowded points by performing an online mean shift clustering. A novel method called θ_WD is then presented that is able to detect loop-like traces on the fly. The experimental results on the GPS datasets of several elders have show that the θ_WD method is effective and efficient in detecting wandering behaviors, in terms of detection performance (AUC > 0.99, and 90% detection rate with less than 5 % of the false alarm rate), as well as time complexity.

Ping-pong robotics with high-speed vision system

Abstract: The performance of vision-based control is usually limited by the low sampling rate of the visual feedback. We address Ping-Pong robotics as a widely studied example which requires high-speed vision for highly dynamic motion control. In order to detect a flying ball accurately and robustly, a multi-threshold segmentation algorithm is applied in a stereo-vision running at 150Hz. Based on the estimated 3D ball positions, a novel two-phase trajectory prediction is exploited to determine the hitting position. Benefiting from the high-speed visual feedback, the hitting position and thus the motion planning of the manipulator are updated iteratively with decreasing error. Experiments are conducted on a 7 degrees of freedom humanoid robot arm. A successful Ping-Pong playing between the robot arm and human is achieved with a high successful rate of 88%.

H ∞ synchronization of Markovian jump master-slave systems with delay

Abstract: The problem of synchronization is investigated for a class of master-slave systems with time-varying delays and Markovian switching parameters. Using an appropriate Lyapunov-Krasovskii functional, a synchronization law which includes the master-slave parameters is established for designing a mode-dependent output feedback control law in terms of LMIs. A numerical example is given to show the effectiveness of the method.

Leveraging multiple simulators for crossing the reality gap

Abstract: We propose a novel approach for transferring evolved control systems from a simulated environment to a real robot. Multiple dynamic simulation systems are simultaneously employed to provide a valid range of simulation variance that can be exploited to generate robust controllers in a purely virtual environment. These controllers can then be directly transferred to a physical robot.

Activity recognition from a wearable camera

Abstract: This paper proposes a novel activity recognition approach from video data obtained with a wearable camera. The objective is to recognise the user's activities from a tiny front-facing camera embedded in his/her glasses. Our system allows carers to remotely access the current status of a specified person, which can be broadly applied to those living with disabilities including the elderly who require cognitive assistance or guidance for daily activities. We collected, trained and tested our system on videos collected from different environmental settings. Sequences of four basic activities (drinking, walking, going upstairs and downstairs) are tested and evaluated in challenging real-world scenarios. An optical flow procedure is used as our primary feature extraction method, from which we downsize, reformat and classify sequence of activities using k-Nearest Neighbour algorithm (k-NN), LogitBoost (on Decision Stumps) and Support Vector Machine (SVM). We suggest the optimal settings of these classifiers through cross-validations and achieve an accuracy of 54.2% to 71.9%. Further smoothing using Hidden Markov Model (HMM) improves the result to 68.5%-82.1%.

Hand-held monocular SLAM in thermal-infrared

Abstract: Thermal-infrared imagery is relatively robust to many of the failure conditions of visual and laser-based SLAM systems, such as fog, dust and smoke. The ability to use thermal-infrared video for localization is therefore highly appealing for many applications. However, operating in thermal-infrared is beyond the capacity of existing SLAM implementations. This paper presents the first known monocular SLAM system designed and tested for hand-held use in the thermal-infrared modality. The implementation includes a flexible feature detection layer able to achieve robust feature tracking in high-noise, low-texture thermal images. A novel approach for structure initialization is also presented. The system is robust to irregular motion and capable of handling the unique mechanical shutter interruptions common to thermal-infrared cameras. The evaluation demonstrates promising performance of the algorithm in several environments.

Automatic calibration for inertial measurement unit

Abstract: The nine degrees-of-freedom (DOF) inertial measurement units (IMU) are generally composed of three kinds of sensor: accelerometer, gyroscope and magnetometer The calibration of these sensor suites not only requires turn-table or purpose-built fixture, but also entails a complex and laborious procedure in data sampling In this paper, we propose a method to calibrate a 9-DOF IMU by using a set of casually sampled raw sensor measurement Our sampling procedure allows the sensor suite to move by hand and only requires about six minutes of fast and slow arbitrary rotations with intermittent pauses It requires neither the specially-designed fixture and equipment, nor the strict sequences of sampling steps At the core of our method are the techniques of data filtering and a hierarchical scheme for calibration All the raw sensor measurements are preprocessed by a series of band-pass filters before use And our calibration scheme makes use of the gravity and the ambient magnetic field as references, and hierarchically calibrates the sensor model parameters towards the minimization of the mis-alignment, scaling and bias errors Moreover, the calibration steps are formulated as a series of function optimization problems and are solved by an evolutionary algorithm Finally, the performance of our method is experimentally evaluated The results show that our method can effectively calibrate the sensor model parameters from one set of raw sensor measurement, and yield consistent calibration results

Improved variable forgetting factor recursive least square algorithm

Abstract: In this paper an improved variable forgetting factor recursive least square (IVFF-RLS) algorithm is proposed. The forgetting factor is adjusted according to the square of a time-averaging estimate of the autocorrelation of a priori and a posteriori errors. The proposed algorithm has fast convergence, and robustness against variable background noise, near-end signal variations and echo path change. The simulation results indicate the superior performances of IVFF-RLS when compared to the RLS and VFF-RLS algorithms.

Automated cervical cell image segmentation using level set based active contour model

Abstract: In this paper, we propose a method based on level set active contour model to sever the nucleus and cytoplast from the cervical smear image. The region of interest (ROI) which contained a main connected cell region has been separated from the smear image after the coarse segmentation by auto dual-threshold segmentation. In the process of fine segmentation, two independent level set functions based on the Chan-Vese model with intra-region similarity and inter-region diversity have been constructed to approximate the cytoplast and nucleus contours. While there may be more than one connected cell regions in the ROI, a method of main cell body and main cell nucleus contour curve extraction has been proposed. We validate the proposed models by numerical experiment and the results show that by means of the adjustment of weight coefficient λ 1 and λ 2 , most cervical cell image with weak edges can be segmented precisely.

Improvement in real-time obstacle detection system for USV

Abstract: A real-time obstacle detection system with vision sensors has been developed for Unmanned Surface Vehicle (USV). The system enables the detection and localization of multiple obstacles in the range from 30 to 300 meters on the sea surface. Field tests have proven the performance and reliability of the system to be satisfactory. This paper focuses on the improvement made in the real-time obstacle detection system during the last year.

Fusing vision and odometry for accurate indoor robot localization

Abstract: For service robotics, localization is an essential component required in many applications, e.g. indoor robot navigation. Today, accurate localization relies mostly on high-end devices, such as A.R.T. DTrack, VICON systems or laser scanners. These systems are often expensive and, thus, require substantial investments. In this paper, our focus is on the development of a localization method using low-priced devices, such as cameras, while being sufficiently accurate in tracking performance. Vision data contains much information and potentially yields high tracking accuracy. However, due to high computational requirements vision-based localization can only be performed at a low frequency. In order to speed up the visual localization and increase accuracy, we combine vision information with robots odometry using a Kalman-Filter. The resulting approach enables sufficiently accurate tracking performance (errors in the range of few cm) at a frequency of about 35Hz. To evaluate the proposed method, we compare our tracking performance with the high precision A.R.T. DTrack localization as ground truth. The evaluations on real robot show that our low-priced localization approach is competitive for indoor robot localization tasks.

A pose measurement method of a non-cooperative GEO spacecraft based on stereo vision

Abstract: Space robotic system is expected to play an increasingly important role in repairing GEO (geostationary orbit) satellites in the future. To perform the servicing mission, the robotic system is firstly required to approach and dock with the target autonomously, for which the measurement of relative pose is the key. It is a challenging task since the existing GEO satellites are generally non-cooperative, i.e. no artificial mark is mounted to aid the measurement. In this paper, a method based on binocular stereo vision is proposed to estimate the pose of a GEO satellite in the final approach phase. It directly takes the natural circular feature on the GEO satellite as the recognized object. Correspondingly, an image processing and pose measurement algorithm is presented to determine the relative position and orientation of the target. This algorithm provides a closed-form solution using simple mathematics, therefore, it is suitable to space applications where the computation capability of the on-board processor is very limited. In addition, it effectively solves the orientation-duality problem for circular feature, requiring neither specific motions of the camera nor a priori knowledge about the radius of the circle. Computer simulations verify the proposed method.

Experimental comparison of 3D vision sensors for mobile robot localization for industrial application: Stereo-camera and RGB-D sensor

Abstract: While RGB-D sensors are becoming more and more popular in mobile robotics laboratories, they are usually not yet adopted for industrial applications. In fact, in this field, depth measurements are generally acquired by means of laser scanners and, when visual information is needed, by means of stereo-cameras. The aim of this paper is to perform an experimental validation, to compare the performance of a stereo-camera and an RGB-D sensor, in a specific application: mobile robot localization for industrial applications. Experiments are performed exploiting artificial landmarks (defined by a self-similar pattern), placed in known positions in the environment.

A system for hand gesture based signature recognition

Abstract: In this paper, we propose a user authentication system based on hand-gesture signature without the need of any handheld device. The system uses a depth image sensor to locate the fingertip and palm mass-center from detected hand region for trajectory processing. Apart from the positional information, the velocity and acceleration information are included as input features for trajectory matching. The system verification performance is evaluated in terms of equal error rate. In addition, an investigation of fusion at feature level is conducted for possible performance enhancement. Our empirical results show the potential of the proposed bare-hand in-air signature system.

Neural-based navigation of a differential-drive mobile robot

Abstract: This paper presents a new neural network based reactive navigation algorithm for wheeled mobile robots (WMR) in unstructured indoor environments. The mobile robot travels to a pre-defined goal position safely and efficiently without any prior map of the environment. This navigation algorithm is optimized by a user-defined objective function which minimizes the traveled distance to the goal position while avoiding obstacles. The network is trained through off-line learning followed by an on-line learning algorithm with guaranteed convergence. The performance of the proposed algorithm is verified over a variety of real unstructured indoor environments using an autonomous mobile robot platform.

Autonomous control of running takeoff and landing for a fixed-wing unmanned aerial vehicle

Abstract: This paper concerns about automatic flight control strategies for the running takeoff and landing of fixed-wing unmanned aerial vehicles (UAVs). Firstly, a dynamic model of fixed-wing UAVs is developed. Then the takeoff and landing control strategies are designed. The flight control strategies are designed to be segmented. A fixed-wing UAV named “Petrel” is used as an experimental platform. Finally, the flight experimental results are proposed to validate the automatic running takeoff and landing controller.

Discriminative dictionary learning for EEG signal classification in Brain-computer interface

Abstract: Modern Brain-computer interface (BCI) technique is essentially based on the classification of the brain signals. The sparse representation classification (SRC) method has been studied for classifying EEG signals of the motor imagery based BCI. The dictionary used in the SRC method is the simple combination of feature vectors which are extracted from the EEG signal-trials by common spatial pattern (CSP) algorithm. In this paper, we propose a method to learn a new dictionary with smaller size and more discriminative ability for the classification. The proposed method, discriminative dictionary learning (DDL), is based on minimizing an objective function containing a reconstructive term and a discriminative term. We apply an iterative scheme to the optimization and transform it to a series of mixed l1-l2 optimizations, which are solved based on separable surrogate functions (SSF) technique. We evaluate the proposed method using the dataset from BCI competition III. The experimental results show that the proposed method outperforms the SRC method.

Distributed output regulation design for multi-agent systems in output-feedback form

Abstract: A distributed output regulation problem is studied for a class of nonlinearly coupled multi-agent systems, taking a decentralized output-feedback form. By a networked internal model with a preassigned topology and a decentralized stabilization design, a distributed regulator is constructed and solves the problem. As an illustration, for a three-mass-three-spring system, a problem of synchronizing every agent motion to a harmonic oscillator can be solved.

Feature extraction through Binary Pattern of Phase Congruency for facial expression recognition

Abstract: Although facial expression plays an important role in human interaction, automated facial expression analysis is still a challenging task. This paper presents a novel facial descriptor based on Phase Congruency (PC) and Local Binary Pattern (LBP) for facial expression recognition. The proposed descriptor, named Binary Pattern of Phase Congruency (BPPC), is an oriented and multi-scale local descriptor that is able to encode various patterns of face images. It is constructed by applying LBP on the oriented PC images. We evaluated the proposed method using the Cohn-Kanade (CK+) database. In our experiment, we achieved an overall detection rate of 93.83% for the six basic emotions. This shows the effectiveness of the proposed method.

Adaptive consensus tracking control of uncertain nonlinear systems: A first-order example

Abstract: In this paper, we consider the problem of designing distributed adaptive consensus tracking controllers for multiple nonlinear systems with unknown parameters and external disturbances. The desired trajectory is time varying given by the state of a reference system, which is only available to a portion of the group of the systems. Besides, the dynamics of the reference state is bounded but unknown to all of the systems. The communication graph characterizing the interactions among the systems is assumed to have undirected, fixed and connected topology. By introducing distributed estimators for the bound of the reference dynamics, two control schemes are proposed to address the problem. In the first scheme, a sign function is employed and perfect consensus tracking can be achieved. In the second scheme, an alternative control law is developed and the chattering phenomenon caused by the sign function can be reduced. However, new challenge will be triggered which is to compensate for possible destabilizing effects of the coupling elements relating to local parameter estimation errors and the synchronization errors of the neighbors. The overall communication graph is firstly reduced to an undirected spanning tree with single system notified of the reference state. Based on this, new synchronization error for each subsystem is then defined as the weighted distance relative to only one of its neighbors. It is shown that all the synchronization errors will converge to a prescribed bound which can be made as small as desired in this case.

An efficient shape analysis method for shrimp quality evaluation

Abstract: Two grading criteria used in determining shrimp product quality and value by the shrimp industry are: 1. Presence or percentage of black spot, measured as a percentage of the total body surface. 2. Shape quality referring to whole shrimp and broken pieces. Black spots (melanoma) on the shrimp surface are evidence of aging shrimp and are considered defects that must be removed from the main production line. Shape quality is measured as the size and the completeness of the body. Broken shrimp pieces are considered a product defect and also must be removed from the main production line. Black spot detection is a simple task for a well-designed machine vision system, which provides consistent and controlled lighting. Shape analysis, on the other hand, is a challenging task because it involves contour extraction and shape analysis. In this paper, a simple, fast, and accurate shape analysis method using Turn Angle Cross-correlation is developed for shrimp quality evaluation. Our analysis results validate that the performance of the proposed shape analysis method is suitable for real-time inspection for commercial applications.

Passivity-based consensus and passification for a class of stochastic multi-agent systems with switching topology

Abstract: This paper studies the passivity-based consensus analysis and the consensus synthesis problem (called passification) for a class of stochastic multi-agent systems subject to external disturbances. Based on Lyapunov methods, graph theory, and slack matrix methods such as the free-weighting matrix and Jensen's integral inequality, a new storage Lyapunov functional is proposed to derive delay-dependent sufficient conditions on mean-square exponential consensus and stochastic passivity for the stochastic multi-agent systems. By proposing passive time-varying stochastic consensus protocols, the solvability conditions for the passification problem are derived based on linearization techniques. A numerical example is provided to illustrate the effectiveness of the theoretical results.

Global robust output regulation for nonlinear multi-agent systems in strict feedback form

Abstract: In this paper, the cooperative global robust output regulation problem for a class of strict feedback nonlinear uncertain multi-agent systems is studied. We first introduce a type of distributed internal model that converts the cooperative global robust output regulation problem into a global robust stabilization problem of a so-called augmented multi-agent system which is in block lower triangular form. We then further present a set of sufficient conditions under which the augmented multi-agent system is globally stabilizable by a distributed state feedback control law.

Stockpile modelling using mobile laser scanner for quality grade control in stockpile management

Abstract: The delivery of a relatively constant quality grade of iron ore is crucial to Australia mining. As a buffer, blending and target grading system, stockpile plays a key role in the iron ore quality control. However, the shape of the stockpile, the quantity housed within the stockpile and the quality grade inside the stockpile are currently unavailable to operators. Therefore, this paper describes a method to generate 3D model of a stockpile with multiple layers using laser to keep tracking the shape changes caused by the stacking and reclaiming operations. Using this 3D model, the volume can be easily calculated anytime. Experiments conducted in laboratory environment indicate good results and proved that the 3D modelling method is accurate and efficient.

Leader-following rendezvous with connectivity preservation of single-integrator multi-agent systems

Abstract: This paper studies the problem of leader-following rendezvous with connectivity preservation for a linear multi-agent system where the leader system is a linear autonomous system and the follower system is a multiple single-integrator system. We propose a distributed state feedback control protocol that is able to maintain the connectivity of the system and, at the same time, achieve asymptotic tracking of all followers to the output of the leader system.

Extending Bayesian RFS SLAM to multi-vehicle SLAM

Abstract: In this paper we present a novel solution to the Multi-Vehicle SLAM (MVSLAM) problem by extending the random finite set (RFS) based SLAM filter framework using two recently developed multi-sensor information fusion approaches. Our solution is based on the modelling of the measurements and the landmark map as RFSs and factorizing the MVSLAM posterior into a product of the joint vehicle trajectories posterior and the landmark map posterior conditioned the vehicle trajectories. The joint vehicle trajectories posterior is propagated using a particle filter while the landmark map posterior conditioned on the vehicle trajectories is propagated using a Gaussian Mixture (GM) implementation of the probability hypothesis density (PHD) filter.

A new method for occupancy grid maps merging: Application to multi-vehicle cooperative local mapping and moving object detection in outdoor environment

Abstract: Autonomous mapping, especially in the form of SLAM (Simultaneous Localization And Mapping), has long since been used for many indoor robotic applications and is also useful in outdoor intelligent vehicle applications such as object detection. Most existing research works on environment mapping and object detection in outdoor applications have been dedicated to single vehicle system. On the other hand, multi-vehicle cooperative perception based on inter-vehicle data sharing can bring considerable benefits in many scenarios that are challenging for a single vehicle system. In this paper, a new method for occupancy grid maps merging is proposed: an objective function based on occupancy likelihood is introduced to measure the consistency degree of maps alignment; genetic algorithm implemented in a dynamic scheme is adopted to optimize the objective function. A scheme of multi-vehicle cooperative local mapping and moving object detection using the proposed occupancy grid maps merging method is also introduced. Real-data tests are given to demonstrate the effectiveness of the introduced method.