Albert-Jan Baerveldt

An Agricultural Mobile Robot with Vision-Based Perception for Mechanical Weed Control

This paper presents an autonomous agricultural mobile robot for mechanical weed control in outdoor environments. The robot employs two vision systems: one gray-level vision system that is able to recognize the row structure formed by the crops and to guide the robot along the rows and a second, color-based vision system that is able to identify a single crop among weed plants. This vision system controls a weeding-tool that removes the weed within the row of crops. The row-recognition system is based on a novel algorithm and has been tested extensively in outdoor field tests and proven to be able to guide the robot with an accuracy of ±2 cm. It has been shown that color vision is feasible for single plant identification, i.e., discriminating between crops and weeds. The system as a whole has been verified, showing that the subsystems are able to work together effectively. A first trial in a greenhouse showed that the robot is able to manage weed control within a row of crops.

A low-cost and low-weight attitude estimation system for an autonomous helicopter

In this paper a low-cost and low-weight attitude estimation system for an autonomous helicopter is presented. The system is based on an inclinometer and a rate gyro. The data coming from the sensors is fused through a complementary filter. In this way the slow dynamics of the inclinometer can be effectively compensated. Tests have shown that a very effective attitude estimation system can be achieved.

Robot localization based on scan-matching—estimating the covariance matrix for the IDC algorithm

We have previously presented a new scan-matching algorithm based on the IDC (iterative dual correspondence) algorithm, which showed a good localization performance even in environments with severe changes. The problem of the IDC algorithm is that there is no good way to estimate a covariance matrix of the position estimate, which prohibits an effective fusion with other position estimates of other sensors. This paper presents two new ways to estimate the covariance matrix. The first estimates the covariance matrix from the Hessian matrix of the error function minimized by the scan-matching algorithm. The second one, which is an off-line method, estimates the covariance matrix of a specific scan, from a specific position by simulating and matching scans around the position. Simulation results show that the covariance matrix provided by the off-line method fully corresponds with the real one. Some preliminary tests on real data indicate that the off-line method gives a good quality value of a specific scan position, which is of great value in map building.

Localization in changing environments - estimation of a covariance matrix for the IDC algorithm

We (1999) have previously presented a new scan-matching algorithm based on the iterative dual correspondence (IDC) algorithm, which showed a good localization performance even in the case of severe changes in the environment. The problem with the IDC algorithm is that there is no good way to estimate the covariance matrix of the position estimate, thus prohibits an effective fusion with other position estimates from other sensors, e.g., by means of the Kalman filter. In this paper we present a new way to estimate the covariance matrix by estimating the Hessian matrix of the error function that is minimized by the IDC scan-matching algorithm. Simulation results show that the estimated covariance matrix correspond well to the real one.

Finger-force measurement-device for hand rehabilitation

The purpose was to develop an extension fingerforce measurement device, and investigate the intraindividual repeatability. The design of the measuring device allows single finger force and whole hand measurements, and the repeatability error on extension finger forces was measured, both on the whole hand, as well as on individual fingers. The tests showed that a repeatability error of less then 15 % can be achieved for single finger measurements and less then 21 % for whole hand measurements.

A vision system for object verification and localization based on local features

An object verification and localization system should answer the question whether an expected object is present in an image or not, i.e. verification, and if present where it is located. Such a sys ...

Modular Neural Network and Classical Reinforcement Learning for Autonomous Robot Navigation: Inhibiting Undesirable Behaviors

Classical reinforcement learning mechanisms and a modular neural network are unified for conceiving an intelligent autonomous system for mobile robot navigation. The conception aims at inhibiting two common navigation deficiencies: generation of unsuitable cyclic trajectories and ineffectiveness in risky configurations. Distinct design apparatuses are considered for tackling these navigation difficulties, for instance: 1) neuron parameter for memorizing neuron activities (also functioning as a learning factor), 2) reinforcement learning mechanisms for adjusting neuron parameters (not only synapse weights), and 3) a inner-triggered reinforcement. Simulation results show that the proposed system circumvents difficulties caused by specific environment configurations, improving the relation between collisions and captures.

Vision-guided mobile robots for design competitions

We present the Teaching system integration in engineering curricula at universities via popular and effective robot-design competitions. In this article, we first present the robot kit we use and discuss the experiences we gained in, and shortcomings of, our robot competitions. We then present the low-cost color vision system, developed especially for our course. We also discuss our experiences in the 1998 competition, where we first made use of the system, and the 2001 competition. Finally, we present the results of a questionnaire given to students who have completed the course.

A low-cost colour vision-system for robot design competitions

In this paper we present a low-cost colour vision system mainly intended for robot design competitions, which nowadays is a popular, project-oriented, way of teaching mechatronics in engineering cu ...

Visual guidance of mobile robots using a neural network

In this paper we present a self-learning method for low-level navigation for autonomous mobile robots, based on a neural network. Both corridor following and obstacle avoidance in indoor environmen ...