Localization Based on Natural Features Detector for Steep Slope Vineyards

Mendes, Jorge; Santos, Filipe Neves dos; Ferraz, Nuno; Couto, Pedro; Morais, Raul

doi:10.1007/s10846-017-0770-8

Cited by 21 publications

(7 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The vision sensor can be used to provide complementary information to exclude the LiDAR points of no interest from data processing. For example, image features were used to separate the LiDAR points of the trunk from other objects in vineyards so that crop rows could be detected correctly [101]. The LiDAR sensor also can be used for obstacle avoidance, but can falsely detect grass, weed, and plant leaves as obstacles, so using vision sensors can help identify real obstacles.…”

Section: Lidar-based Navigationmentioning

confidence: 99%

A Review of High-Throughput Field Phenotyping Systems: Focusing on Ground Robots

2022

Plant Phenomics

View full text Add to dashboard Cite

Manual assessments of plant phenotypes in the field can be labor-intensive and inefficient. The high-throughput field phenotyping systems and in particular robotic systems play an important role to automate data collection and to measure novel and fine-scale phenotypic traits that were previously unattainable by humans. The main goal of this paper is to review the state-of-the-art of high-throughput field phenotyping systems with a focus on autonomous ground robotic systems. This paper first provides a brief review of nonautonomous ground phenotyping systems including tractors, manually pushed or motorized carts, gantries, and cable-driven systems. Then, a detailed review of autonomous ground phenotyping robots is provided with regard to the robot’s main components, including mobile platforms, sensors, manipulators, computing units, and software. It also reviews the navigation algorithms and simulation tools developed for phenotyping robots and the applications of phenotyping robots in measuring plant phenotypic traits and collecting phenotyping datasets. At the end of the review, this paper discusses current major challenges and future research directions.

show abstract

Section: Lidar-based Navigationmentioning

confidence: 99%

A Review of High-Throughput Field Phenotyping Systems: Focusing on Ground Robots

2022

Plant Phenomics

View full text Add to dashboard Cite

show abstract

“…The green bounding boxes represent the ground truth, and the red ones, the detections. Figures [10][11][12] represent the precision × recall curves p(r), for all the considered configurations: the three retrained models, either with the IoU threshold t equals to 0.5 and 0.65, under the three evaluation sets. Table 3 summarizes the AP for all the configurations.…”

Section: A Detection Performancementioning

confidence: 99%

“…The vine trunks can be used as landmarks for the SLAM, and to build a vineyard map. There are solutions to perform these tasks using range sensors [5], [9] or camera systems [10], [11], based on traditional methods such as Kalman Filters (KF), image processing, and others. However, to the best of our knowledge, the use of DL [12] to detect vine trunks is still nonexistent in the literature.…”

Section: Introductionmentioning

confidence: 99%

Visual Trunk Detection Using Transfer Learning and a Deep Learning-Based Coprocessor

et al. 2020

Self Cite

View full text Add to dashboard Cite

Agricultural robotics is nowadays a complex, challenging, and exciting research topic. Some agricultural environments present harsh conditions to robotics operability. In the case of steep slope vineyards, there are several challenges: terrain irregularities, characteristics of illumination, and inaccuracy/unavailability of signals emitted by the Global Navigation Satellite System (GNSS). Under these conditions, robotics navigation becomes a challenging task. To perform these tasks safely and accurately, the extraction of reliable features or landmarks from the surrounding environment is crucial. This work intends to solve this issue, performing accurate, cheap, and fast landmark extraction in steep slope vineyard context. To do so, we used a single camera and an Edge Tensor Processing Unit (TPU) provided by Google's USB Accelerator as a small, high-performance, and low power unit suitable for image classification, object detection, and semantic segmentation. The proposed approach performs object detection using Deep Learning (DL)-based Neural Network (NN) models on this device to detect vine trunks. To train the models, Transfer Learning (TL) is used on several pre-trained versions of MobileNet V1 and MobileNet V2. A benchmark between the two models and the different pre-trained versions is performed. The models are pre-trained in a built in-house dataset, that is publicly available containing 336 different images with approximately 1,600 annotated vine trunks. There are considered two vineyards, one using camera images with the conventional infrared filter and others with an infrablue filter. Results show that this configuration allows a fast vine trunk detection, with MobileNet V2 being the most accurate retrained detector, achieving an overall Average Precision of 52.98%. We briefly compare the proposed approach with the state-of-the-art Tiny YOLO-V3 running on Jetson TX2, showing the outperformance of the adopted system in this work. Additionally, it is also shown that the proposed detectors are suitable for the Localization and Mapping problems.INDEX TERMS Deep learning, transfer learning, convolutional neural networks, tensor processing unit.

show abstract

“…3b). The AgRob v16 robot [11] is a modular mobile robot designed to navigate and manipulate inside Douro's vineyard, sustaining its complexities. At the rear of the robot, there is a lightweight manipulator, the Robotis Manipulator-H, with about 6 kg and a maximum payload of 3 kg.…”

Section: Fig 2 Vine Model To Algorithms Benchmarkingmentioning

confidence: 99%

“…So, naturally, and to overcome these production limitations, viticulturists need and desire new robotic solutions to automate agricultural tasks (monitoring, pruning, spraying, and harvesting). The special Douro's steep slope vineyard features bring a high number of robotic challenges which needs to overcome, namely, robot self-localisation, robot environment perception, environment modelling and decision-making, to reach a fully autonomous and intelligent system [11,15]. The resolution of these robotic challenges, besides improve the robotics technology in agriculture and viticulture, they will reduce the requirement of labour work in the most basic agricultural tasks, such as watering, pruning, harvesting or monitoring.…”

Section: Introductionmentioning

confidence: 99%

Path Planning Algorithms Benchmarking for Grapevines Pruning and Monitoring

Magalhães

Santos

Martins

et al. 2019

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Labour shortage is a reality in agriculture. Farmers are asking for solutions to automate agronomic tasks, such as monitoring, pruning, spraying, and harvesting. The automation of these tasks requires, most of the time, the use of robotic arms to mimic human arms capabilities. The current robotic arm based solutions available, both in the market and in the scientific sphere, have several limitations, such as, lowspeed manipulation, the path planning algorithms are not aware of the requirements of the agricultural tasks (robotic motion and manipulation synchronisation), and require active perception tuning to the end-target point. This work benchmarks algorithms from open manipulation planning library (OMPL) considering a cost-effective six-degree freedom manipulator in a simulated vineyard. The OMPL planners shown a very low performance under demanding pruning tasks. The best and most promising results are performed and obtained by BiTRRT. However, further work is needed to increase its performance and reduce planning time. This benchmark work helps the reader to understand the limitations of each algorithm and when to use them.

show abstract

Localization Based on Natural Features Detector for Steep Slope Vineyards

Cited by 21 publications

References 17 publications

A Review of High-Throughput Field Phenotyping Systems: Focusing on Ground Robots

A Review of High-Throughput Field Phenotyping Systems: Focusing on Ground Robots

Visual Trunk Detection Using Transfer Learning and a Deep Learning-Based Coprocessor

Path Planning Algorithms Benchmarking for Grapevines Pruning and Monitoring

Contact Info

Product

Resources

About