Vineyard yield estimation based on the analysis of high resolution images obtained with artificial illumination at night.

This paper presents a method for vineyard yield estimation based on the analysis of high-resolution images obtained with artificial illumination at night. First, this paper assesses different pixel-based segmentation methods in order to detect reddish grapes: threshold based, Mahalanobis distance, Bayesian classifier, linear color model segmentation and histogram segmentation, in order to obtain the best estimation of the area of the clusters of grapes in this illumination conditions. The color spaces tested were the original RGB and the Hue-Saturation-Value (HSV). The best segmentation method in the case of a non-occluded reddish table-grape variety was the threshold segmentation applied to the H layer, with an estimation error in the area of 13.55%, improved up to 10.01% by morphological filtering. Secondly, after segmentation, two procedures for yield estimation based on a previous calibration procedure have been proposed: (1) the number of pixels corresponding to a cluster of grapes is computed and converted directly into a yield estimate; and (2) the area of a cluster of grapes is converted into a volume by means of a solid of revolution, and this volume is converted into a yield estimate; the yield errors obtained were 16% and -17%, respectively.

Ambient intelligence application based on environmental measurements performed with an assistant mobile robot.

This paper proposes the use of an autonomous assistant mobile robot in order to monitor the environmental conditions of a large indoor area and develop an ambient intelligence application. The mobile robot uses single high performance embedded sensors in order to collect and geo-reference environmental information such as ambient temperature, air velocity and orientation and gas concentration. The data collected with the assistant mobile robot is analyzed in order to detect unusual measurements or discrepancies and develop focused corrective ambient actions. This paper shows an example of the measurements performed in a research facility which have enabled the detection and location of an uncomfortable temperature profile inside an office of the research facility. The ambient intelligent application has been developed by performing some localized ambient measurements that have been analyzed in order to propose some ambient actuations to correct the uncomfortable temperature profile.

A proposal for automatic fruit harvesting by combining a low cost stereovision camera and a robotic arm.

This paper proposes the development of an automatic fruit harvesting system by combining a low cost stereovision camera and a robotic arm placed in the gripper tool. The stereovision camera is used to estimate the size, distance and position of the fruits whereas the robotic arm is used to mechanically pickup the fruits. The low cost stereovision system has been tested in laboratory conditions with a reference small object, an apple and a pear at 10 different intermediate distances from the camera. The average distance error was from 4% to 5%, and the average diameter error was up to 30% in the case of a small object and in a range from 2% to 6% in the case of a pear and an apple. The stereovision system has been attached to the gripper tool in order to obtain relative distance, orientation and size of the fruit. The harvesting stage requires the initial fruit location, the computation of the inverse kinematics of the robotic arm in order to place the gripper tool in front of the fruit, and a final pickup approach by iteratively adjusting the vertical and horizontal position of the gripper tool in a closed visual loop. The complete system has been tested in controlled laboratory conditions with uniform illumination applied to the fruits. As a future work, this system will be tested and improved in conventional outdoor farming conditions.

An embedded real-time red peach detection system based on an OV7670 camera, ARM cortex-M4 processor and 3D look-up tables.

This work proposes the development of an embedded real-time fruit detection system for future automatic fruit harvesting. The proposed embedded system is based on an ARM Cortex-M4 (STM32F407VGT6) processor and an Omnivision OV7670 color camera. The future goal of this embedded vision system will be to control a robotized arm to automatically select and pick some fruit directly from the tree. The complete embedded system has been designed to be placed directly in the gripper tool of the future robotized harvesting arm. The embedded system will be able to perform real-time fruit detection and tracking by using a three-dimensional look-up-table (LUT) defined in the RGB color space and optimized for fruit picking. Additionally, two different methodologies for creating optimized 3D LUTs based on existing linear color models and fruit histograms were implemented in this work and compared for the case of red peaches. The resulting system is able to acquire general and zoomed orchard images and to update the relative tracking information of a red peach in the tree ten times per second.

Definition of linear color models in the RGB vector color space to detect red peaches in orchard images taken under natural illumination.

This work proposes the detection of red peaches in orchard images based on the definition of different linear color models in the RGB vector color space. The classification and segmentation of the pixels of the image is then performed by comparing the color distance from each pixel to the different previously defined linear color models. The methodology proposed has been tested with images obtained in a real orchard under natural light. The peach variety in the orchard was the paraguayo (Prunus persica var. platycarpa) peach with red skin. The segmentation results showed that the area of the red peaches in the images was detected with an average error of 11.6%; 19.7% in the case of bright illumination; 8.2% in the case of low illumination; 8.6% for occlusion up to 33%; 12.2% in the case of occlusion between 34 and 66%; and 23% for occlusion above 66%. Finally, a methodology was proposed to estimate the diameter of the fruits based on an ellipsoidal fitting. A first diameter was obtained by using all the contour pixels and a second diameter was obtained by rejecting some pixels of the contour. This approach enables a rough estimate of the fruit occlusion percentage range by comparing the two diameter estimates.

Two-dimensional radial laser scanning for circular marker detection and external mobile robot tracking.

This paper presents the use of an external fixed two-dimensional laser scanner to detect cylindrical targets attached to moving devices, such as a mobile robot. This proposal is based on the detection of circular markers in the raw data provided by the laser scanner by applying an algorithm for outlier avoidance and a least-squares circular fitting. Some experiments have been developed to empirically validate the proposal with different cylindrical targets in order to estimate the location and tracking errors achieved, which are generally less than 20 mm in the area covered by the laser sensor. As a result of the validation experiments, several error maps have been obtained in order to give an estimate of the uncertainty of any location computed. This proposal has been validated with a medium-sized mobile robot with an attached cylindrical target (diameter 200 mm). The trajectory of the mobile robot was estimated with an average location error of less than 15 mm, and the real location error in each individual circular fitting was similar to the error estimated with the obtained error maps. The radial area covered in this validation experiment was up to 10 m, a value that depends on the radius of the cylindrical target and the radial density of the distance range points provided by the laser scanner but this area can be increased by combining the information of additional external laser scanners.

Characterization of a low-cost optical flow sensor when using an external laser as a direct illumination source.

In this paper, a low cost optical flow sensor is combined with an external laser device to measure surface displacements and mechanical oscillations. The measurement system is based on applying coherent light to a diffuser surface and using an optical flow sensor to analyze the reflected and transferred light to estimate the displacement of the surface or the laser spot. This work is focused on the characterization of this measurement system, which can have the optical flow sensor placed at different angles and distances from the diffuser surface. The results have shown that the displacement of the diffuser surface is badly estimated when the optical mouse sensor is placed in front of the diffuser surface (angular orientation >150°) while the highest sensitivity is obtained when the sensor is located behind the diffuser surface and on the axis of the laser source (angular orientation 0°). In this case, the coefficient of determination of the measured displacement, R(2), was very high (>0.99) with a relative error of less than 1.29%. Increasing the distance between the surface and the sensor also increased the sensitivity which increases linearly, R(2) = 0.99. Finally, this measurement setup was proposed to measure very low frequency mechanical oscillations applied to the laser device, up to 0.01 Hz in this work. The results have shown that increasing the distance between the surface and the optical flow sensor also increases the sensitivity and the measurement range.

Design and implementation of a biomimetic turtle hydrofoil for an autonomous underwater vehicle.

This paper presents the design and implementation of a turtle hydrofoil for an Autonomous Underwater Vehicle (AUV). The final design of the AUV must have navigation performance like a turtle, which has also been the biomimetic inspiration for the design of the hydrofoil and propulsion system. The hydrofoil design is based on a National Advisory Committee for Aeronautics (NACA) 0014 hydrodynamic profile. During the design stage, four different propulsion systems were compared in terms of propulsion path, compactness, sealing and required power. The final implementation is based on a ball-and-socket mechanism because it is very compact and provides three degrees of freedom (DoF) to the hydrofoil with very few restrictions on the propulsion path. The propulsion obtained with the final implementation of the hydrofoil has been empirically evaluated in a water channel comparing different motion strategies. The results obtained have confirmed that the proposed turtle hydrofoil controlled with a mechanism with three DoF generates can be used in the future implementation of the planned AUV.