Branch detection for apple trees trained in fruiting wall architecture using depth features and Regions-Convolutional Neural Network (R-CNN)

Zhang, Jing; He, Long; Karkee, Manoj; Zhang, Qin; Xin, Zhang; Gao, Zongmei

doi:10.1016/j.compag.2018.10.029

Cited by 105 publications

(52 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, end-to-end training has been conducted on popular architectures in a supervised manner by streamlining the training procedure. The common architectures are CNN and recurrent neural network (RNN) [30,31]. CNN has been widely used for image analysis, and RNN is becoming more and more popular.…”

Section: Neural Networkmentioning

confidence: 99%

Deep Learning Application in Plant Stress Imaging: A Review

et al. 2020

Self Cite

View full text Add to dashboard Cite

Plant stress is one of major issues that cause significant economic loss for growers. The labor-intensive conventional methods for identifying the stressed plants constrain their applications. To address this issue, rapid methods are in urgent needs. Developments of advanced sensing and machine learning techniques trigger revolutions for precision agriculture based on deep learning and big data. In this paper, we reviewed the latest deep learning approaches pertinent to the image analysis of crop stress diagnosis. We compiled the current sensor tools and deep learning principles involved in plant stress phenotyping. In addition, we reviewed a variety of deep learning applications/functions with plant stress imaging, including classification, object detection, and segmentation, of which are closely intertwined. Furthermore, we summarized and discussed the current challenges and future development avenues in plant phenotyping.

show abstract

Section: Neural Networkmentioning

confidence: 99%

Deep Learning Application in Plant Stress Imaging: A Review

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recent work has focused on convolutional neural networks. From the use of networks such as R-CNN [34], [35] and OverFeat [36], the fruits are classified and bounding box regression by proposing object. Later work such as Faster R-CNN proposed the Region Proposal Network (RPN) for the generation of informative regions, which aims to automatically extract image features and quickly and accurately detect corn seedlings at different growth stages in complex field environments [37].…”

Section: B Object Detectionmentioning

confidence: 99%

Self-Supervised Collaborative Multi-Network for Fine-Grained Visual Categorization of Tomato Diseases

Yang

Chen

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Artificial recognition of tomato diseases is often time-consuming, laborious and subjective. For tomato disease images, it is difficult to find small discriminative features between different tomato diseases, which can bring challenges to fine-grained visual categorization of tomato leaf-based images. Therefore, we propose a novel model, which consists of 3 networks, including a Location network, a Feedback network, and a Classification network, named LFC-Net. At the same time, a self-supervision mechanism is proposed in the model, which can effectively detect informative regions of tomato image without the need for manual annotation such as bounding boxes/parts. Based on the consideration of the consistency between category of the image and informativeness of the image, we design a novel training paradigm. The Location network of the model first detects informative regions in the tomato image, and optimizes iterations under the guidance of the Feedback network. Then, the Classification network uses informative regions proposed by the Location network and the full image of the tomato for classification. Our model can be regarded as a multi-network collaboration, and networks can progress together. Compared with the pre-trained model on ImageNet, our model achieves the most advanced performance in the tomato dataset, with accuracy up to 99.7%. This work demonstrates that our model has a high accuracy and has the potential to be applied to other vegetable and fruit datasets, which can provide a reference for the prevention and control of tomato diseases. INDEX TERMS Fine-grained visual categorization, multi-network, self-supervised, tomato diseases This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.

show abstract

“…Tao et al (2017) used support vector machines and a genetic algorithm to classify apple tree canopy images into fruit, branches and leaves. Similarly, Zhang et al (2018b) used color and depth information collected by a Microsoft Kinect sensor to train a faster R-CNN, which showed better accuracy in classifying branches, trunks and background compared to the same without depth information.…”

Section: Fig 146 a Robotic Harvesting System Developed At Washingtomentioning

confidence: 99%

The use of agricultural robots in orchard management

Zhang¹,

Karkee²,

Tabb³

2019

Burleigh Dodds Series in Agricultural Science

Self Cite

View full text Add to dashboard Cite

The use of robotic or automated machines in orchard operations is associated primarily with insufficient labor availability and rapidly increasing labor costs in tree fruit production and is critical for improving yield of high-quality fruit with minimal dependence on seasonal human labor. Primarily, mechanized or robotic orchard management operations (after establishment of the trees) include pruning, thinning, spraying and harvesting.Pruning is an operation to grow fruit trees into a desired shape for improving sunlight penetration into the canopies, supporting more effective (e.g. more human-and machinefriendly) orchard operations, and regulating location and amount of fruiting sites that can lead to improved yield and quality of fruit. Pruning is physically cutting and removing parts of a tree, such as cutting branches back and/or removing small limbs entirely following some guidelines. This operation requires workers have a specific level of knowledge and skill to perform the job well. Pruning is performed both in the winter (also called "dormant pruning") and in the summer. Currently, research and development of robotic pruning is focused primarily on dormant pruning, therefore, we will focus on dormant pruning for the remainder of this chapter.Fruit trees' bloom often includes many more flowers than are required for producing desired yield and quality of a commercial crop. Crop thinning is a practice to manage/control crop load by removing a portion of the bloom and/or young fruit on the tree either selectively or non-selectively to improve the size and quality of the remaining fruit. Thinning is commonly performed during and shortly after the bloom period to reduce the number of set fruit using either chemical thinners or physical means. However, thinning can occur later in the season to remove poor quality fruit or those growing too close together to ensure good quality and size of the remaining ones using only physical means. Currently research and development of robotic thinning technologies is more focused on blossom thinning, and therefore, we will focus on this aspect of thinning in this chapter.Pest control is an essential operation in orchard management to protect the yield and quality of fruit from various types of pests. One solution to achieve effective pest control is to spray recommended pesticides at the correct time and correct rate to the target area. Due to the confined environment in tree fruit orchards, spraying pesticides is inherently hazardous for the operators driving a sprayer. Robotic sprayers are probably the best solution to ensure operator safety and efficiency in pest control.Fruit harvesting for fresh market consumption requires a large seasonal semi-skilled workforce which has created a critical labor-related risk of not having enough workers to perform this time-sensitive task. In addition, harvesting labor is already the most significant

show abstract

Branch detection for apple trees trained in fruiting wall architecture using depth features and Regions-Convolutional Neural Network (R-CNN)

Cited by 105 publications

References 17 publications

Deep Learning Application in Plant Stress Imaging: A Review

Deep Learning Application in Plant Stress Imaging: A Review

Self-Supervised Collaborative Multi-Network for Fine-Grained Visual Categorization of Tomato Diseases

The use of agricultural robots in orchard management

Contact Info

Product

Resources

About