Detection and Characterization of Cherries: A Deep Learning Usability Case Study in Chile

Villacrés, Juan; Cheein, Fernando Auat

doi:10.3390/agronomy10060835

Cited by 42 publications

(18 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Regarding the precision in the identification of objects in digital images, Reis et al [32] achieves a precision of 97% in the identification of red grapes using a pixel identification method similar to this work. Villacrés and Auat [4] were able to identify cherries with a precision of 85% using faster R-CNN. Si et al [13] identified apples using stereoscopic vision with a precision of 89.5%.…”

Section: Discussionmentioning

confidence: 99%

“…Other works in the literature, besides identification, use the count of objects as productivity parameter [4,14,[40][41][42][43]. As explained before, counting of objects is not a precise indicator for productivity, and that is one reason this works introduces the PI.…”

Section: Discussionmentioning

confidence: 99%

“…Some further control of the environment can also be done computationally with a pre-processing step where targets are first identified before performing a classification at the pixel level. The identification of tomatoes can be done with a Convolutional Neural Network (CNN) trained to identify tomatoes [4,37]. A preselection of targets with Faster R-CNN focuses the PI calculation only on tomatoes, thus avoiding any foreign object that may introduce errors in the calculation of the PI.…”

Section: Pre-processingmentioning

confidence: 99%

“…The application of technology in food production ranges from the use of IoT (Internet of Things) [2] through automatic control based on wireless sensors [3] to cutting-edge arti-ficial intelligence techniques [4]. Currently, among the most popular artificial intelligence techniques is deep learning [5], where a perceptron [6] allows the creation of neural networks that facilitate the classification and identification of patterns.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Definition and Application of a Computational Parameter for the Quantitative Production of Hydroponic Tomatoes Based on Artificial Neural Networks and Digital Image Processing

et al. 2021

View full text Add to dashboard Cite

This work presents an alternative method, referred to as Productivity Index or PI, to quantify the production of hydroponic tomatoes using computer vision and neural networks, in contrast to other well-known metrics, such as weight and count. This new method also allows the automation of processes, such as tracking of tomato growth and quality control. To compute the PI, a series of computational processes are conducted to calculate the total pixel area of the displayed tomatoes and obtain a quantitative indicator of hydroponic crop production. Using the PI, it was possible to identify objects belonging to hydroponic tomatoes with an error rate of 1.07%. After the neural networks were trained, the PI was applied to a full crop season of hydroponic tomatoes to show the potential of the PI to monitor the growth and maturation of tomatoes using different dosages of nutrients. With the help of the PI, it was observed that a nutrient dosage diluted with 50% water shows no difference in yield when compared with the use of the same nutrient with no dilution.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Pre-processingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Definition and Application of a Computational Parameter for the Quantitative Production of Hydroponic Tomatoes Based on Artificial Neural Networks and Digital Image Processing

et al. 2021

View full text Add to dashboard Cite

show abstract

“…A total of 4427 RGB images and 512 HS images were obtained. In terms of sampling, only at most 122 RGB images were collected in [ 25 , 46 ] which used machine learning methods: Faster R-CNN and random forest, respectively, while 15,000 images were acquired in [ 47 ], which also used Faster R-CNN. [ 30 , 48 , 49 ] collected and used 557, 300, and 240 HSI images for classification using deep learning methods: ResNeXt, GAN, and AlexNet, respectively.…”

Section: Methodsmentioning

confidence: 99%

Multimodal Deep Learning and Visible-Light and Hyperspectral Imaging for Fruit Maturity Estimation

Garillos-Manliguez

Chiang

2021

Sensors

View full text Add to dashboard Cite

Fruit maturity is a critical factor in the supply chain, consumer preference, and agriculture industry. Most classification methods on fruit maturity identify only two classes: ripe and unripe, but this paper estimates six maturity stages of papaya fruit. Deep learning architectures have gained respect and brought breakthroughs in unimodal processing. This paper suggests a novel non-destructive and multimodal classification using deep convolutional neural networks that estimate fruit maturity by feature concatenation of data acquired from two imaging modes: visible-light and hyperspectral imaging systems. Morphological changes in the sample fruits can be easily measured with RGB images, while spectral signatures that provide high sensitivity and high correlation with the internal properties of fruits can be extracted from hyperspectral images with wavelength range in between 400 nm and 900 nm—factors that must be considered when building a model. This study further modified the architectures: AlexNet, VGG16, VGG19, ResNet50, ResNeXt50, MobileNet, and MobileNetV2 to utilize multimodal data cubes composed of RGB and hyperspectral data for sensitivity analyses. These multimodal variants can achieve up to 0.90 F1 scores and 1.45% top-2 error rate for the classification of six stages. Overall, taking advantage of multimodal input coupled with powerful deep convolutional neural network models can classify fruit maturity even at refined levels of six stages. This indicates that multimodal deep learning architectures and multimodal imaging have great potential for real-time in-field fruit maturity estimation that can help estimate optimal harvest time and other in-field industrial applications.

show abstract

Potato crop stress identification in aerial images using deep learning‐based object detection

et al. 2021

View full text Add to dashboard Cite

Recent research on the application of remote sensing and deep learning-based analysis in precision agriculture demonstrated a potential for improved crop management and reduced environmental impacts of agricultural production. Despite the promising results, the practical relevance of these technologies for field deployment requires novel algorithms that are customized for analysis of agricultural images and robust to implementation on natural field imagery. The paper presents an approach for analyzing aerial images of a potato (Solanum tuberosum L.) crop using deep neural networks. The main objective is to demonstrate automated spatial recognition of healthy vs. stressed crop at a plant level. Specifically, we examine premature plant senescence resulting in drought stress on 'Russet Burbank' potato plants. We propose a novel deep learning (DL) model for detecting crop stress, named Retina-UNet-Ag. The proposed architecture is a variant of Retina-UNet and includes connections from low-level semantic representation maps to the feature pyramid network. The paper also introduces a dataset of aerial field images acquired with a Parrot Sequoia camera. The dataset includes manually annotated bounding boxes of healthy and stressed plant regions. Experimental validation demonstrated the ability for distinguishing healthy and stressed plants in field images, achieving an average dice score coefficient (DSC) of 0.74. A comparison to related state-of-the-art DL models for object detection revealed that the presented approach is effective for this task. The proposed method is conducive toward the assessment and recognition of potato crop stress in aerial field images collected under natural conditions.

show abstract

Detection and Characterization of Cherries: A Deep Learning Usability Case Study in Chile

Cited by 42 publications

References 28 publications

Definition and Application of a Computational Parameter for the Quantitative Production of Hydroponic Tomatoes Based on Artificial Neural Networks and Digital Image Processing

Definition and Application of a Computational Parameter for the Quantitative Production of Hydroponic Tomatoes Based on Artificial Neural Networks and Digital Image Processing

Multimodal Deep Learning and Visible-Light and Hyperspectral Imaging for Fruit Maturity Estimation

Potato crop stress identification in aerial images using deep learning‐based object detection

Contact Info

Product

Resources

About