J.R. Martinez-de Dios

A cooperative perception system for multiple UAVs: Application to automatic detection of forest fires

This paper presents a cooperative perception system for multiple heterogeneous unmanned aerial vehicles (UAVs). It considers different kind of sensors: infrared and visual cameras and fire detectors. The system is based on a set of multipurpose low-level image-processing functions including segmentation, stabilization of sequences of images, and geo-referencing, and it also involves data fusion algorithms for cooperative perception. It has been tested in field experiments that pursued autonomous multi-UAV cooperative detection, monitoring, and measurement of forest fires. This paper presents the overall architecture of the perception system, describes some of the implemented cooperative perception techniques, and shows experimental results on automatic forest fire detection and localization with cooperating UAVs.

Cooperative Fire Detection using Unmanned Aerial Vehicles

The paper presents a framework for cooperative fire detection by means of a fleet of heterogeneous UAVs. Computer vision techniques are used to detect and localize fires from infrared and visual images and other data provided by the cameras and other sensors on-board the UAVs. The paper deals with the techniques used to decrease the uncertainty in fire detection and increase the accuracy in fire localisation by means of the cooperation of the information provided by several UAVs. The presented methods have been developed in the COMETS multi-UAV project.

Computer vision techniques for forest fire perception

This paper presents computer vision techniques for forest fire perception involving measurement of forest fire properties (fire front, flame height, flame inclination angle, fire base width) required for the implementation of advanced forest fire-fighting strategies. The system computes a 3D perception model of the fire and could also be used for visualizing the fire evolution in remote computer systems. The presented system integrates the processing of images from visual and infrared cameras. It applies sensor fusion techniques involving also telemetry sensors, and GPS. The paper also includes some results of forest fire experiments.

Computer vision and robotics techniques in fish farms

This paper presents new low-cost systems for the automation of some fish farm operations. Particularly, computer vision is applied to non-contact fish weight estimation. Stereo vision systems with synchronised convergent cameras are employed to perform fish 3-D segmentation in tanks and sea cages. Several pre-processing algorithms are applied to compensate for illumination local variations. The approach applied for fish 3-D segmentation consists in detecting in both images certain fish features. Once these points have been detected and validated in both images, the fish are 3-D segmented by applying stereo vision matching considerations. Fish weight is estimated by using simple length-weight relations well known in the aquaculture domain. The paper also briefly describes robotics systems for fish feeding and underwater pond cleaning, which can be also used to implement the above mentioned computer vision techniques for the fish estimation.

Cooperation Between UAS and Wireless Sensor Networks for Efficient Data Collection in Large Environments

This paper describes a method for collection of data from Wireless Sensor Network (WSN) deployed in large environments using Unmanned Aerial Systems (UAS). Unlike existing approaches, in which the WSN and the UAS act as independent units, the main novelty of the proposed method is that UAS and WSN cooperate to increase the performance of the mission. The proposed method presents two main cooperative behaviors: (1) the results of the WSN operation are used to update the UAS flight plan and; (2) the UAS trajectory is considered in the operation of the WSN in order to improve the data collection performance. The proposed method outperforms non-cooperative UAS-based collection approaches and traditional ground multi-hop collection schemes. The method has been experimented in the airfield of Bellavista in Seville (Spain) in March 2011.

Data Retrieving From Heterogeneous Wireless Sensor Network Nodes Using UAVs

This paper describes a method and experimental results of a flight planning method that takes into account uncertainties to determine a safe UAV trajectory. It uses particle filters to predict UAV trajectories taking into account the model of the UAV and of the atmospheric conditions and also considering uncertainties. A waypoint generation module computes intermediate waypoints in order to ensure that the trajectory achieves the required levels of safety (avoids forbidden zones) and mission achievement (passes through way-zones). The method has been applied to collection of data from wireless sensor network and has been validated in the airfield of Bollullos in the Spanish province of Seville.

Cooperative localization and tracking with a camera-based WSN

This paper presents a vision-based system for cooperative object detection, localization and tracking using Wireless Sensor Networks (WSNs). The proposed system exploits the distributed sensing capabilities, communication infrastructure and parallel computing capabilities of the WSN. To reduce the bandwidth requirements, the images captured are processed at each camera node with the objective of extracting the location of the object on each image plane, which is transmitted to the WSN. The measures from all the camera nodes are processed by means of sensor fusion techniques such as Maximum Likelihood (ML) and Extended Kalman Filter (EKF). The paper describes hardware and software aspects and presents some experimental results.

Efficient robot-sensor network distributed SEIF range-only SLAM

This paper is motivated by schemes of robot-sensor network cooperation where sensor nodes (beacons) are used as landmarks for Range-Only (RO) Simultaneous Localization and Mapping (SLAM). Most existing RO-SLAM techniques consider beacons as passive devices disregarding the sensing, computing and communication capabilities they are actually endowed with. This paper proposes a Range-Only scheme based on Sparse Extended Information Filters (SEIF) that efficiently exploits their capabilities. The robot computes the SLAM prediction stage and distributes the update stage among beacons within its sensing area. The proposed scheme naturally integrates robot-beacon and inter-beacon measurements, significantly improving map and also robot estimations. Our scheme inherits from SEIF its efficiency and scalability and further reduces robot computational burden by exploiting the beacons computing capability. As a result, it has lower error and lower computer requirements than traditional methods. This paper presents the scheme, evaluates and compares its performance in simulations and real experiments.

A Multiresolution Threshold Selection Method Based on Training

This paper presents a new training-based threshold selection method for grey level images. One of the main limitations of existing threshold selection methods is the lack of capacity of adaptation to specific vision applications. The proposed method represents a procedure to adapt threshold selection methods to specific applications. The proposed method is based on the analysis of multiresolution decompositions of the image histogram, which is supervised by fuzzy systems in which the particularities of the specific applications were introduced. The method has been extensively applied in various computer vision applications, one of which is described in this paper.

A Real-Time Image Stabilization System Based on Fourier-Mellin Transform

The paper presents a robust real-time image stabilization system based on the Fourier-Mellin transform. The system is capable of performing image capture-stabilization-display at a rate of standard video on a general Pentium III at 800 MHz without any specialized hardware and the use of any particular software platforms. This paper describes the theoretical basis of the image matching used and the practical aspects considered to increase its robustness and accuracy as well as the optimizations carried out for its real-time implementation. The system has been submitted to extensive practical experimentation in several applications showing high robustness.