Multi-sensor Fusion Scheme Research Articles

Runway Extraction in Low Visibility Conditions Based on Sensor Fusion Method

This paper presents the design and implementation of a low-cost vision-based system for an aircraft during approach and landing under low visibility. The system is based on combination of multisensor fusion strategy and image processing algorithm, where no expensive equipments, such as millimeter wave radar, are needed. It is aiming to alleviate restrictions in airspace and airport capacity in low visibility conditions by augmenting the naturally existing visual cues. With the fusion approach, pilots are able to identify the runway under low visibility conditions.

Sensor Fusion Method for Horizon Detection From an Aircraft in Low Visibility Conditions

The approach and landing of an aircraft under low visibility has been a subject of great interest in recent years. Although much advanced equipment has been designed and used on airplanes to improve the safety threshold in manipulating it, low visibility has remained the potential threat of causing controlled flight into terrain and runway intrusion. Besides, low visibility is also the main cause for flight delay, which reduces the efficiency of transportation systems. Clear landmarks on the ground will significantly improve pilots' spatial awareness under low visibility conditions, and among all landmarks, horizon position is the key one to count on for successfully controlling an aircraft during landing. The purpose of this paper is to design a horizon detection method based on a multisensor fusion strategy suitable for poor visibility conditions. With the fusion approach, pilots are able to recover/reconstruct the horizon under low visibility conditions. Analysis of experimental results and comparisons with reported methods are furnished at the end of this paper.

A Comprehensive Approach for On-line Tool Condition Identification in Metal Cutting Processes

Abstract—On-line cutting tool condition monitoring becomes one of the most critical requirements in machining processes for improving the efficiency and the autonomy of CNC machine tools. The processes can be significantly improved by using an intelligent integration of sensor information to detect and identify accurately the tool condition under various cutting parameters. This paper presents a structured and comprehensive approach for on-line tool condition identification in metal cutting processes using ANN based multi-sensor fusion strategy. Various sensing techniques are combined with different preprocessing techniques to select suitable monitoring indices and then numerous models for on-line tool condition identification. The proposed approach is built progressively by examining monitoring indices from various aspects and making modeling decision step by step. The results indicate a significant improvement and a good reliability in identifying various tool conditions regardless of the variation in cutting parameters.

Research on the Strategy of Multi Sensor Information Fusion in Coal Mine Disaster Scene

Different types of disasters occur frequently in coal mines. This paper analyzed the characteristics of different disasters, chosen the corresponding sensors to collect the information of disaster scene, and discussed the methods of multi sensor information fusion. Lastly, the multi-sensor information fusion strategies for fire, gas outburst, flood, and roof collapse were proposed in this research.

Cutting Tool Condition Monitoring in Machining Processes - A Comprehensive Approach Using ANN Based Multisensor Fusion Strategy

On-line cutting tool condition monitoring becomes one of the most critical requirements in machining processes for improving the efficiency and the autonomy of CNC machine tools. The processes can be significantly improved by using an intelligent integration of sensor information to detect and identify accurately the tool condition under various cutting parameters. This paper presents a structured and comprehensive approach for tool condition monitoring in machining processes using ANN based multisensor fusion strategy. Various sensing techniques are combined to select suitable monitoring indices and several models are proposed to establish the relationship between tool condition and the selected monitoring indices. The proposed approach is built progressively by examining monitoring indices from various aspects and making monitoring decision step by step. The results indicate a significant improvement and a good reliability in identifying various tool conditions regardless of the variation in cutting parameters.

Multisensor data fusion for impact-echo testing of concrete structures

Based on the investigation of spatial variations of impact-echo signals for different source and receiver locations, a simple multisensor data fusion strategy is proposed to increase the accuracy of nondestructive evaluation of concrete structures using the impact-echo test. The data fusion strategy fuses the ratios between spectral amplitudes of the delamination and concrete bottom echo signals (D/B ratios) from multiple source–receiver arrays. The fused D/B ratios demonstrate different characteristics for test locations above the delamination, above the sound concrete, and across the delamination boundary. These characteristics can be used to accurately locate the delamination and its boundaries without increasing testing time. The applicability of the multisensor data fusion for impact-echo testing is validated using both numerical simulation and experimental testings.

Wearable Sensor-Based Hand Gesture and Daily Activity Recognition for Robot-Assisted Living

In this paper, we address natural human-robot interaction (HRI) in a smart assisted living (SAIL) system for the elderly and the disabled. Two common HRI problems are studied: hand gesture recognition and daily activity recognition. For hand gesture recognition, we implemented a neural network for gesture spotting and a hierarchical hidden Markov model for context-based recognition. For daily activity recognition, a multisensor fusion scheme is developed to process motion data collected from the foot and the waist of a human subject. Experiments using a prototype wearable sensor system show the effectiveness and accuracy of our algorithms.

Analysis of Multiresolution-Based Fusion Strategies for a Dual Infrared System

A dual infrared system to assist a driver in bad visibility conditions is studied. The problem of selecting the best multiresolution-based image fusion technique is addressed with reference to automotive scenarios. A new method for objective evaluation of multisensor image fusion strategies is presented for the optimal design of the fusion process. Multiresolution-based fusion methodologies are compared, and experimental results obtained from a prototype dual infrared camera system are shown and analyzed. Numerical results, in terms of the quality of the fused images and of the computational load, are presented and discussed. The effectiveness of the dual infrared system in urban and extraurban automotive scenarios is illustrated with a number of examples.

A Road Matching Method for Precise Vehicle Localization Using Hybrid Bayesian Network

This article presents a multisensor fusion strategy for a novel road-matching method designed to support real-time navigational features within advanced driver assistance systems. In road navigation, context, integrity, reliability and accuracy are essential qualities for road-matching methods. Particularly, managing multihypotheses is a useful strategy to treat ambiguous situations in the road-matching task. In this study, multisensor fusion and multimodal estimation are realized using a hybrid Bayesian network. To manage multihypothesis, multimodal estimation is proposed. Experimental results, using data from antilock braking system sensors, a differential global positioning system receiver, and an accurate digital roadmap illustrate the performance of the proposed approach, especially in ambiguous situations.

A region-based multi-sensor image fusion scheme using pulse-coupled neural network

For most image fusion algorithms split relationship among pixels and treat them more or less independently, this paper proposes a region-based image fusion scheme using pulse-coupled neural network (PCNN), which combines aspects of feature and pixel-level fusion. The basic idea is to segment all different input images by PCNN and to use this segmentation to guide the fusion process. In order to determine PCNN parameters adaptively, this paper brings forward an adaptive segmentation algorithm based on a modified PCNN with the multi-thresholds determined by a novel water region area method. Experimental results demonstrate that the proposed fusion scheme has extensive application scope and it outperforms the multi-scale decomposition based fusion approaches, both in visual effect and objective evaluation criteria, particularly when there is movement in the objects or mis-registration of the source images.

A fuzzy modeling approach to decision fusion under uncertainty

A multi-sensor decision fusion scheme is presented in which the probabilities associated with the local sensor decisions are known to vary in a nonrandom fashion around their design values. The uncertainties associated with the local decisions are modeled by means of fuzzy sets. A Bayesian approach is used to design the optimum fusion rule for the case where the local sensor decisions are statistically independent across the sensors. In order to reach a crisp decision, the global Bayesian risk is defuzzified using a criterion for mapping fuzzy sets on to the real line. The performance of the optimum fusion rule obtained is illustrated by means of a numerical example.

Multisensor-based robotic manipulation in an uncalibrated manufacturing workcell

This paper deals with the problem of locating a part moving on a rotating disc conveyor with the aid of a single CCD camera and conveyor encoders and subsequently manipulating the part with a robot manipulator. The precise position and orientation of the disc conveyor with respect to the robot manipulator is assumed to be a priori unknown. Likewise the position and orientation of the camera with respect to the robot is also assumed to be a priori unknown. In order to manipulate the rotating part, the robot controller relies on data obtained from multiple sensors. Among the various sensors that have been used in this paper, we have a single CCD camera which is held fixed in the workspace of the robot and which is able to detect features of the part as it rotates with the conveyor. Likewise, the camera is also able to detect features of the end-effector as it moves in the workspace. Additionally, the conveyor is assumed to be equipped with an encoder that measures the angular position of the rotating disc. What makes the robotic manipulation problem hard is that the precise position of the camera and the robot are a priori assumed to be unknown. Thus, the path that the end effector needs to follow is not a priori known in the coordinates of the robot but needs to be computed during the course of the manipulation. In this paper, a new multisensor fusion scheme is proposed that performs this computation. The advantage of the scheme is that it greatly weakens requirement on the image processing speed. Overall, our algorithm is able to handle lack of calibration between the robot, the camera and the conveyor, advantage of which is that it significantly shortens the setup time of the manufacturing workcell. The paper touches upon both the theory and the experiments.

3D Part Manipulation Aided by Uncalibrated Mono-Camera *

Abstract 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 apriori 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 non-planar, i.e. the feature points observed on the part is assumed to be located arbitrarily in IR 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