Instance Segmentation and Localization of Strawberries in Farm Conditions for Automatic Fruit Harvesting

Ge, Yuanyue; Xiong, Ya; From, Pål Johan

doi:10.1016/j.ifacol.2019.12.537

Cited by 39 publications

(19 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…So far, the application of deep learning models in precision agriculture have shown advantages for automatic feature extraction and learning, transfer learning, quick adaptation to a new problem, dealing with heterogeneous big data, and obtaining higher accuracy and excellent performance [31], [32]. Convolutional neural networks (CNN) and their derivatives have shown to be among the most successful techniques in image classification and recognition [24], [33].…”

Section: Related Workmentioning

confidence: 99%

“…The framework yielded a classification accuracy of 82.1%. Ge et al [33] employed the Mask Region-CNN model to detect and classify different ripening levels (raw, pink, and ripe) of strawberries in farm conditions. Mohtar et al [36] adopted an Inception-v3 model to classify six stages of ripening of mangosteen fruit with a classification accuracy of 91.9%.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Detection of Mulberry Ripeness Stages Using Deep Learning Models

et al. 2021

View full text Add to dashboard Cite

Ripeness classification is one of the most challenging tasks in the postharvest management of mulberry fruit. The risks of microbial contamination and human error in manual sorting are significant; it may result in quality degradation and wasting of processed products. Due to advanced developments in computer vision and machine learning, automated sorting became possible. This study presents the results of developing and testing a computer vision-based application using convolutional neural networks (CNNs) for the classification of mulberry fruit ripening stages. To reduce the training cost and improve the accuracy of classification, transfer learning was used to fine-tune the CNN models. The CNN models in the test include DenseNet, Inception-v3, ResNet-18, ResNet-50, and AlexNet. Transfer learning was used to fine-tune the models and improve the accuracy of classification. The AlexNet and ResNet-18 networks exhibited the best performance with 98.32% and 98.65% overall accuracy for classifying the ripeness of white and black mulberries, respectively. Moreover, the performance of the models did not change when the data sets of both genotypes were mixed. The ResNet-18 was able to classify both genotype and ripeness from 600 fruit images in 2.36 min with an overall accuracy of 98.03%, which was superior to other architectures. It indicates that the model could be used for precise classification of the ripening stages of mulberries and other horticultural products, as a part of an automated sorting system.INDEX TERMS Convolutional neural network, computer vision, online detection, ripening classification, transfer learning.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Detection of Mulberry Ripeness Stages Using Deep Learning Models

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Ge et al [85] utilized the ResNet architecture for strawberry detection. Its uses the basic CNN to extract features from the input images.…”

Section: Resnet (Rnet)mentioning

confidence: 99%

Recent Advancements in Fruit Detection and Classification Using Deep Learning Techniques

Ukwuoma

Qin

Heyat

et al. 2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Recent advances in computer vision have allowed broad applications in every area of life, and agriculture is not left out. For the agri-food industry, the use of advanced technology is essential. Owing to deep learning’s capability to learn robust features from images, it has witnessed enormous application in several fields. Fruit detection and classification remains challenging due to the form, color, and texture of different fruit species. While studying the impact of computer vision on fruit detection and classification, we pointed out that till 2018 many conventional machine learning methods were utilized while a few methods exploited the application of deep learning methods for fruit detection and classification. This has prompted us to pursue an extensive study on surveying and implementing deep learning models for fruit detection and classification. In this article, we intensively discussed the datasets used by many scholars, the practical descriptors, the model’s implementation, and the challenges of using deep learning to detect and categorize fruits. Lastly, we summarized the results of different deep learning methods applied in previous studies for the purpose of fruit detection and classification. This review covers the study of recently published articles that utilized deep learning models for fruit identification and classification. Additionally, we also implemented from scratch a deep learning model for fruit classification using the popular dataset “Fruit 360” to make it easier for beginner researchers in the field of agriculture to understand the role of deep learning in the agriculture domain.

show abstract

“…Among the most important factors can be distinguished the way of harvesting-the strawberries must be harvested in such a way as to eliminate or minimize bruising, the damage to the cover and the leakage of fruit juices-as well as the correct identification of the fruit. Ge et al [13] presented the results of research on the use of the Deep Convolutional Neural Network to identify the degree of ripeness and shape deformation of harvested strawberries.…”

Section: Robotic Strawberry Harvestmentioning

confidence: 99%

Analysis of the Pneumatic System Parameters of the Suction Cup Integrated with the Head for Harvesting Strawberry Fruit

Kurpaska

Sobol

Pedryc

et al. 2020

Sensors

View full text Add to dashboard Cite

Fruit and vegetable harvest efficiency depends on the mechanization and automation of production. The available literature lacks the results of research on the applicability of pneumatic end effectors among grippers for the robotic harvesting of strawberries. To determine their practical applications, a series of tests was performed. They included the determination of the morphological indicators of the strawberry, fruit suction force, the real stress exerted by fruit suckers and the degree of fruit damage. The fruits’ morphological indicators included the relationships between the weight and geometrical dimensions of the tested fruit, the equivalent diameter, and the sphericity coefficient. The fruit suction force was determined on a stand equipped with a vacuum pump, and control and measurement instruments, as well as a MTS 2 testing machine. The necrosis caused by tissue damage to the fruits by suction cup adhesion was assessed by counting the necrosis surface areas using the LabView programme. The assessment of the necrosis was conducted immediately upon the test’s performance, after 24 and after 72h. The stress values were calculated by referring the values of the suction forces obtained to the surface of the suction cup face. The tests were carried out with three constructions of suction cups and three positions of suction cup faces on the fruits’ surface. The research shows that there is a possibility for using pneumatic suction cups in robotic picking heads. The experiments performed indicate that the types of suction cups constructions, and the zones and directions of the suction cups’ application to the fruit significantly affect the values of the suction forces and stresses affecting the fruit. The surface areas of the necrosis formed depend mainly on the time that elapses between the test and their assessment. The weight of strawberry fruit in the conducted experiment constituted from 13.6% to 23.1% of the average suction force.

show abstract

Instance Segmentation and Localization of Strawberries in Farm Conditions for Automatic Fruit Harvesting

Cited by 39 publications

References 15 publications

Detection of Mulberry Ripeness Stages Using Deep Learning Models

Detection of Mulberry Ripeness Stages Using Deep Learning Models

Recent Advancements in Fruit Detection and Classification Using Deep Learning Techniques

Analysis of the Pneumatic System Parameters of the Suction Cup Integrated with the Head for Harvesting Strawberry Fruit

Contact Info

Product

Resources

About