In-Field Automatic Detection of Grape Bunches under a Totally Uncontrolled Environment

Ghiani, Luca; Sassu, Alberto; Palumbo, Francesca; Mercenaro, Luca; Gambella, Filippo

doi:10.3390/s21113908

Cited by 24 publications

(20 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The lack of readily available valuable data for training and validation purposes is the key obstacle to adopting machine learning procedures in the horticulture industry for the aim of predicting fruit quality [ 32 ]. Thus, the fresh grape samples for the required data were collected from seven-year-old Flame Seedless grapevines, which were cultivated in private vineyards located at QoraKhargin, Gharb El-Nobarya region, Beheira Governorate, Egypt (30°34′55.3″ N, 29°53′25.4″ E) during the 2021 growing season.…”

Section: Methodsmentioning

confidence: 99%

Physico-Chemical Properties Prediction of Flame Seedless Grape Berries Using an Artificial Neural Network Model

Al-Saif

Abdel-Sattar

Aboukarima

et al. 2022

Foods

View full text Add to dashboard Cite

The grape is a very well-liked fruit that is valued for its distinct flavor and several health benefits, including antioxidants, anthocyanins, soluble sugars, minerals, phenolics, flavonoids, organic acids, and vitamins, which significantly improve the product’s overall quality. Today’s supply chain as a whole needs quick and easy methods for evaluating fruit quality. Thus, the objective of this study was to estimate the quality attributes of Flame Seedless grape berries cultivated under various agronomical management and other practices using color space coordinates (berry L*, berry a*, and berry b*) as inputs in an artificial neural network (ANN) model with the best topology of (3-20-11). Satisfactory predictions based on the R2 range, which was 0.9817 to 0.9983, were obtained for physical properties (i.e., berry weight, berry length, and berry diameter as well as berry adherence strength) and chemical properties (i.e., anthocyanin, total soluble solids (TSS), TSS/titratable acidity, total sugars, titratable acidity, reducing sugars, and non-reducing sugars). Meanwhile, we also performed a contribution analysis to analyze the relative importance of CIELab colorimeter parameters of berries L*, a*, and b* to determine the main fruit quality. In terms of relative contribution, berry b* contributed relatively largely to berry weight, berry adherence strength, TSS, TSS/titratable acidity, titratable acidity, total sugars, reducing sugars, and non-reducing sugars and a* contributed relatively largely to anthocyanin, berry length, and berry diameter. The developed ANN prediction model can aid growers in enhancing the quality of Flame Seedless grape berries by selecting suitable agronomical management and other practices to avoid potential quality issues that could affect consumers of them. This research demonstrated how color space coordinates and ANN model may well be utilized to evaluate the Flame seedless grape berries’ quality.

show abstract

Section: Methodsmentioning

confidence: 99%

Physico-Chemical Properties Prediction of Flame Seedless Grape Berries Using an Artificial Neural Network Model

Al-Saif

Abdel-Sattar

Aboukarima

et al. 2022

Foods

View full text Add to dashboard Cite

show abstract

“…Lottes et al [30] used an encoder-decoder Fully Convolutional Network (FCN) model to identify crops and weeds during field operations. Ghiani et al [31] used Mask R-CNN with ResNet101 which was trained with the dataset COCO, as a backbone for detecting grape branches in the tree.…”

Section: Segmentation By Cnnmentioning

confidence: 99%

“…Ghiani et al [31] used Mask R-CNN with ResNet101 as a backbone which was pretrained with the dataset COCO (https://cocodataset.org/#home, accessed on 18 November 2021) for detecting grape branches on the tree. An open-source dataset GrapeCS-ML [73] containing more than 2000 images without annotation of fifteen grape varieties at different stages of development in three Australian vineyards was used to train the model.…”

Section: Fruit Detection and Yield Forecastmentioning

confidence: 99%

“…In addition, agricultural image datasets are more complex due to outdoor conditions, the object of interest usually occupying a very small and uncentered part of the image, similarity between objects and background, blocking the object with leaves and branches, multiple objects in one image, and many more [31,63,65,68,79,82]. To be applicable in the real world, the dataset should fully represent the state of the environment.…”

Section: Disease Detectionmentioning

confidence: 99%

See 1 more Smart Citation

A Review of the Challenges of Using Deep Learning Algorithms to Support Decision-Making in Agricultural Activities

et al. 2022

View full text Add to dashboard Cite

Deep Learning has been successfully applied to image recognition, speech recognition, and natural language processing in recent years. Therefore, there has been an incentive to apply it in other fields as well. The field of agriculture is one of the most important fields in which the application of deep learning still needs to be explored, as it has a direct impact on human well-being. In particular, there is a need to explore how deep learning models can be used as a tool for optimal planting, land use, yield improvement, production/disease/pest control, and other activities. The vast amount of data received from sensors in smart farms makes it possible to use deep learning as a model for decision-making in this field. In agriculture, no two environments are exactly alike, which makes testing, validating, and successfully implementing such technologies much more complex than in most other industries. This paper reviews some recent scientific developments in the field of deep learning that have been applied to agriculture, and highlights some challenges and potential solutions using deep learning algorithms in agriculture. The results in this paper indicate that by employing new methods from deep learning, higher performance in terms of accuracy and lower inference time can be achieved, and the models can be made useful in real-world applications. Finally, some opportunities for future research in this area are suggested.

show abstract

“…The best-performing model achieved an MAE % of 11.79. In Ghiani et al [14], the Mask R-CNN with ResNet101 as a backbone was trained for detecting grape branches on the tree. The model achieved an mAP of 92.78%.…”

Section: Introductionmentioning

confidence: 99%

Real-Time Detection of Vine Trunk for Robot Localization Using Deep Learning Models Developed for Edge TPU Devices

et al. 2022

View full text Add to dashboard Cite

The concept of the Internet of Things (IoT) in agriculture is associated with the use of high-tech devices such as robots and sensors that are interconnected to assess or monitor conditions on a particular plot of land and then deploy the various factors of production such as seeds, fertilizer, water, etc., accordingly. Vine trunk detection can help create an accurate map of the vineyard that the agricultural robot can rely on to safely navigate and perform a variety of agricultural tasks such as harvesting, pruning, etc. In this work, the state-of-the-art single-shot multibox detector (SSD) with MobileDet Edge TPU and MobileNet Edge TPU models as the backbone was used to detect the tree trunks in the vineyard. Compared to the SSD with MobileNet-V1, MobileNet-V2, and MobileDet as backbone, the SSD with MobileNet Edge TPU was more accurate in inference on the Raspberrypi, with almost the same inference time on the TPU. The SSD with MobileDet Edge TPU achieved the second-best accurate model. Additionally, this work examines the effects of some features, including the size of the input model, the quantity of training data, and the diversity of the training dataset. Increasing the size of the input model and the training dataset increased the performance of the model.

show abstract

In-Field Automatic Detection of Grape Bunches under a Totally Uncontrolled Environment

Cited by 24 publications

References 49 publications

Physico-Chemical Properties Prediction of Flame Seedless Grape Berries Using an Artificial Neural Network Model

Physico-Chemical Properties Prediction of Flame Seedless Grape Berries Using an Artificial Neural Network Model

A Review of the Challenges of Using Deep Learning Algorithms to Support Decision-Making in Agricultural Activities

Real-Time Detection of Vine Trunk for Robot Localization Using Deep Learning Models Developed for Edge TPU Devices

Contact Info

Product

Resources

About