Sandro Augusto Magalhães

Abstract
Several thousand grapevine varieties exist, with even more naming identifiers. Adequate specialized labor is not available for proper classification or identification of grapevines, making the value of commercial vines uncertain. Traditional methods, such as genetic analysis or ampelometry, are time-consuming, expensive, and often require expert skills that are even rarer. New vision-based systems benefit from advanced and innovative technology and can be used by nonexperts in ampelometry. To this end, deep learning (DL) and machine learning (ML) approaches have been successfully applied for classification purposes. This work extends the state of the art by applying digital ampelometry techniques to larger grapevine varieties. We benchmarked MobileNet v2, ResNet-34, and VGG-11-BN DL classifiers to assess their ability for digital ampelography. In our experiment, all the models could identify the vines' varieties through the leaf with a weighted F1 score higher than 92%.

Abstract
Purpose: Visual perception enables robots to perceive the environment. Visual data is processed using computer vision algorithms that are usually time-expensive and require powerful devices to process the visual data in real-time, which is unfeasible for open-field robots with limited energy. This work benchmarks the performance of different heterogeneous platforms for object detection in real-time. This research benchmarks three architectures: embedded GPU-Graphical Processing Units (such as NVIDIA Jetson Nano 2 GB and 4 GB, and NVIDIA Jetson TX2), TPU-Tensor Processing Unit (such as Coral Dev Board TPU), and DPU-Deep Learning Processor Unit (such as in AMD-Xilinx ZCU104 Development Board, and AMD-Xilinx Kria KV260 Starter Kit). Methods: The authors used the RetinaNet ResNet-50 fine-tuned using the natural VineSet dataset. After the trained model was converted and compiled for target-specific hardware formats to improve the execution efficiency.Conclusions and Results: The platforms were assessed in terms of performance of the evaluation metrics and efficiency (time of inference). Graphical Processing Units (GPUs) were the slowest devices, running at 3 FPS to 5 FPS, and Field Programmable Gate Arrays (FPGAs) were the fastest devices, running at 14 FPS to 25 FPS. The efficiency of the Tensor Processing Unit (TPU) is irrelevant and similar to NVIDIA Jetson TX2. TPU and GPU are the most power-efficient, consuming about 5 W. The performance differences, in the evaluation metrics, across devices are irrelevant and have an F1 of about 70 % and mean Average Precision (mAP) of about 60 %.

Abstract
The efficiency of agricultural practices depends on the timing of their execution. Environmental conditions, such as rainfall, and crop-related traits, such as plant phenology, determine the success of practices such as irrigation. Moreover, plant phenology, the seasonal timing of biological events (e.g., cotyledon emergence), is strongly influenced by genetic, environmental, and management conditions. Therefore, assessing the timing the of crops' phenological events and their spatiotemporal variability can improve decision making, allowing the thorough planning and timely execution of agricultural operations. Conventional techniques for crop phenology monitoring, such as field observations, can be prone to error, labour-intensive, and inefficient, particularly for crops with rapid growth and not very defined phenophases, such as vegetable crops. Thus, developing an accurate phenology monitoring system for vegetable crops is an important step towards sustainable practices. This paper evaluates the ability of computer vision (CV) techniques coupled with deep learning (DL) (CV_DL) as tools for the dynamic phenological classification of multiple vegetable crops at the subfield level, i.e., within the plot. Three DL models from the Single Shot Multibox Detector (SSD) architecture (SSD Inception v2, SSD MobileNet v2, and SSD ResNet 50) and one from You Only Look Once (YOLO) architecture (YOLO v4) were benchmarked through a custom dataset containing images of eight vegetable crops between emergence and harvest. The proposed benchmark includes the individual pairing of each model with the images of each crop. On average, YOLO v4 performed better than the SSD models, reaching an F1-Score of 85.5%, a mean average precision of 79.9%, and a balanced accuracy of 87.0%. In addition, YOLO v4 was tested with all available data approaching a real mixed cropping system. Hence, the same model can classify multiple vegetable crops across the growing season, allowing the accurate mapping of phenological dynamics. This study is the first to evaluate the potential of CV_DL for vegetable crops' phenological research, a pivotal step towards automating decision support systems for precision horticulture.

Abstract
The world wine sector is a multi-billion dollar industry with a wide range of economic activities. Therefore, it becomes crucial to monitor the grapevine because it allows a more accurate estimation of the yield and ensures a high-quality end product. The most common way of monitoring the grapevine is through the leaves (preventive way) since the leaves first manifest biophysical lesions. However, this does not exclude the possibility of biophysical lesions manifesting in the grape berries. Thus, this work presents three pre-trained YOLO models (YOLOv5x6, YOLOv7-E6E, and YOLOR-CSP-X) to detect and classify grape bunches as healthy or damaged by the number of berries with biophysical lesions. Two datasets were created and made publicly available with original images and manual annotations to identify the complexity between detection (bunches) and classification (healthy or damaged) tasks. The datasets use the same 10,010 images with different classes. The Grapevine Bunch Detection Dataset uses the Bunch class, and The Grapevine Bunch Condition Detection Dataset uses the OptimalBunch and DamagedBunch classes. Regarding the three models trained for grape bunches detection, they obtained promising results, highlighting YOLOv7 with 77% of mAP and 94% of the F1-score. In the case of the task of detection and identification of the state of grape bunches, the three models obtained similar results, with YOLOv5 achieving the best ones with an mAP of 72% and an F1-score of 92%.

Abstract
The increasing world population, growing need for agricultural products, and labour shortages have driven the growth of robotics in agriculture. Tasks such as fruit harvesting require extensive hours of work during harvest periods and can be physically exhausting. Autonomous robots bring more efficiency to agricultural tasks with the possibility of working continuously. This paper proposes a stackable 3 DoF SCARA manipulator for tomato harvesting. The manipulator uses a custom electronic circuit to control DC motors with an endless gear at each joint and uses a camera and a Tensor Processing Unit (TPU) for fruit detection. Cascaded PID controllers are used to control the joints with magnetic encoders for rotational feedback, and a time-of-flight sensor for prismatic movement feedback. Tomatoes are detected using an algorithm that finds regions of interest with the red colour present and sends these regions of interest to an image classifier that evaluates whether or not a tomato is present. With this, the system calculates the position of the tomato using stereo vision obtained from a monocular camera combined with the prismatic movement of the manipulator. As a result, the manipulator was able to position itself very close to the target in less than 3 seconds, where an end-effector could adjust its position for the picking.