Carlota Salinas

Combination of RGB and Multispectral Imagery for Discrimination of Cabernet Sauvignon Grapevine Elements

This paper proposes a sequential masking algorithm based on the K-means method that combines RGB and multispectral imagery for discrimination of Cabernet Sauvignon grapevine elements in unstructured natural environments, without placing any screen behind the canopy and without any previous preparation of the vineyard. In this way, image pixels are classified into five clusters corresponding to leaves, stems, branches, fruit and background. A custom-made sensory rig that integrates a CCD camera and a servo-controlled filter wheel has been specially designed and manufactured for the acquisition of images during the experimental stage. The proposed algorithm is extremely simple, efficient, and provides a satisfactory rate of classification success. All these features turn out the proposed algorithm into an appropriate candidate to be employed in numerous tasks of the precision viticulture, such as yield estimation, water and nutrients needs estimation, spraying and harvesting.

Multisensory system for fruit harvesting robots. Experimental testing in natural scenarios and with different kinds of crops

The motivation of this research was to explore the feasibility of detecting and locating fruits from different kinds of crops in natural scenarios. To this end, a unique, modular and easily adaptable multisensory system and a set of associated pre-processing algorithms are proposed. The offered multisensory rig combines a high resolution colour camera and a multispectral system for the detection of fruits, as well as for the discrimination of the different elements of the plants, and a Time-Of-Flight (TOF) camera that provides fast acquisition of distances enabling the localisation of the targets in the coordinate space. A controlled lighting system completes the set-up, increasing its flexibility for being used in different working conditions. The pre-processing algorithms designed for the proposed multisensory system include a pixel-based classification algorithm that labels areas of interest that belong to fruits and a registration algorithm that combines the results of the aforementioned classification algorithm with the data provided by the TOF camera for the 3D reconstruction of the desired regions. Several experimental tests have been carried out in outdoors conditions in order to validate the capabilities of the proposed system.

Catadioptric panoramic stereovision for humanoid robots

This paper proposes a novel design of a reconfigurable humanoid robot head, based on biological likeness of human being so that the humanoid robot could agreeably interact with people in various everyday tasks. The proposed humanoid head has a modular and adaptive structural design and is equipped with three main components: frame, neck motion system and omnidirectional stereovision system modules. The omnidirectional stereovision system module being the last module, a motivating contribution with regard to other computer vision systems implemented in former humanoids, it opens new research possibilities for achieving human-like behaviour. A proposal for a real-time catadioptric stereovision system is presented, including stereo geometry for rectifying the system configuration and depth estimation. The methodology for an initial approach for visual servoing tasks is divided into two phases, first related to the robust detection of moving objects, their depth estimation and position calculation, and second the development of attention-based control strategies. Perception capabilities provided allow the extraction of 3D information from a wide range of visions from uncontrolled dynamic environments, and work results are illustrated through a number of experiments.

A New Approach for Combining Time-of-Flight and RGB Cameras Based on Depth-Dependent Planar Projective Transformations

Image registration for sensor fusion is a valuable technique to acquire 3D and colour information for a scene. Nevertheless, this process normally relies on feature-matching techniques, which is a drawback for combining sensors that are not able to deliver common features. The combination of ToF and RGB cameras is an instance that problem. Typically, the fusion of these sensors is based on the extrinsic parameter computation of the coordinate transformation between the two cameras. This leads to a loss of colour information because of the low resolution of the ToF camera, and sophisticated algorithms are required to minimize this issue. This work proposes a method for sensor registration with non-common features and that avoids the loss of colour information. The depth information is used as a virtual feature for estimating a depth-dependent homography lookup table (Hlut). The homographies are computed within sets of ground control points of 104 images. Since the distance from the control points to the ToF camera are known, the working distance of each element on the Hlut is estimated. Finally, two series of experimental tests have been carried out in order to validate the capabilities of the proposed method.

VIS-NIR, SWIR and LWIR Imagery for Estimation of Ground Bearing Capacity

Ground bearing capacity has become a relevant concept for site-specific management that aims to protect soil from the compaction and the rutting produced by the indiscriminate use of agricultural and forestry machines. Nevertheless, commonly known techniques for its estimation are cumbersome and time-consuming. In order to alleviate these difficulties, this paper introduces an innovative sensory system based on Visible-Near InfraRed (VIS-NIR), Short-Wave InfraRed (SWIR) and Long-Wave InfraRed (LWIR) imagery and a sequential algorithm that combines a registration procedure, a multi-class SVM classifier, a K-means clustering and a linear regression for estimating the ground bearing capacity. To evaluate the feasibility and capabilities of the presented approach, several experimental tests were carried out in a sandy-loam terrain. The proposed solution offers notable benefits such as its non-invasiveness to the soil, its spatial coverage without the need for exhaustive manual measurements and its real time operation. Therefore, it can be very useful in decision making processes that tend to reduce ground damage during agricultural and forestry operations.

Validation of a Multisensory System for Fruit Harvesting Robots in Lab Conditions

This paper presents a multisensory system for the detection and localisation of fruits that are candidates to be harvested by a robotic manipulator. The devices that have been selected as primary sensors for this purpose are a high resolution colour camera, a multispectral imaging system that consists of a motorised filter wheel, and a Time-Of-Flight 3D camera. A controlled lighting system completes the set-up. The progressive RGB camera and the multispectral imaging system acquire the basic data inputs for the detection of areas of interest that belong to the fruits, whereas the Time-Of-Flight 3D camera provides fast acquisition of accurate distances enabling the localisation of the targets in the coordinate space. Several experimental tests have been carried out in laboratory conditions in order to evaluate the capabilities of the proposed multisensory system.

A perception system for accurate automatic control of an articulated bus

The authors acknowledge partial funding of this research under: Robocity2030-II S2009/DPI-1559 and IMADE PIE/62/2008 (Comunidad de Madrid, ALDESA CONSTRUCCIONES SA, MAXIMASDE); FORTUNA D/026706/09 (Agencia Espanola de Cooperacion Internacional para el Desarrollo, AECID). Dr. Hector Montes acknowledges support form Universidad Tecnologica de Panama and from CSIC under grant JAE-Doc.

Lateral and longitudinal control system for the automation of an articulated bus

The authors acknowledge partial funding of this research under: Robocity2030 S-2009/DPI-1559, and IMADE PIE/62/2008 (Comunidad de Madrid, ALDESA CONSTRUCCIONES SA, MAXIMASDE); FORTUNA D/026706/09 (Agencia Espanola de Cooperacion Internacional para el Desarrollo, AECID). Dr. Hector Montes acknowledges support form Universidad Tecnologica de Panama and from CSIC under grant JAE-Doc.

Automatic Control of a Large Articulated Vehicle

This work presents the development of the automatic control of an articulated bus. It is describes the most relevant concepts of the architecture of perception and actuation implemented on a large wheeled vehicle, Volvo B10M articulated bus of 18.75m in length. In this work, some experimental results of strategies for lateral and longitudinal control of the bus are shown, when it travels on a guide rail through a guiding mechatronics box. Additionally, this article presents the concepts of the guiding rail detection and obstacle detection by means of two Lidar systems for the test lane detection in advance, and for the pedestrians’ safety, respectively. Several experimental results are presented in this manuscript.

HIGH-ACCURACY MEASUREMENT SYSTEM FOR SECURE STEERING OF LARGE VEHICLES

A measurement system of high accuracy is of great influence for guidance for automated vehicles. This paper presents the description of the high-accuracy measurement system implemented on a large mobile vehicle, an articulated bus Volvo B10M of 18 m in length. Several experiments have been performed on the test lane, with small radii of curvature, where a guide rail has been embedded. The measurement system consists of a guiding mechanical system, which roll within a guide rail designed for this purpose. The measurement system is capable of sending the bus control system the angular position of the guide rail with respect to the current position of the bus. The measurement of the angular position of the guide rail is obtained by means of two incremental encoders with resolution up to 0.036 degrees. In addition, the measurement system acquires the advance speed data from the bus. In this paper several experimental results with the utilization of the measurement system are presented, with the purpose to develop the lateral and longitudinal control algorithms to be implemented in the bus.