Unsupervised Learning Method for Plant and Leaf Segmentation

Al-Vshakarji, Noor M.; Kassim, Yasmin M.; Palaniappan, Kannappan

doi:10.1109/aipr.2017.8457935

Cited by 18 publications

(11 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…well but require a comprehensive set of training data to match the variation expected in real-life data. An unsupervised plant segmentation method that uses k-means clustering with an EM (expectation and maximization) algorithm has also been reported ( Al-Shakarji et al., 2017 ). In addition, deep learning general object detection approaches, such as R-CNN networks ( Huang et al., 2017 ), have been used for disease quantification ( Fuentes et al., 2017 ) and for detection of maize plants in field trials using Lidar-imaging ( Jin et al., 2018 ).…”

Section: Introductionmentioning

confidence: 99%

Greenotyper: Image-Based Plant Phenotyping Using Distributed Computing and Deep Learning

et al. 2020

View full text Add to dashboard Cite

Image-based phenotype data with high temporal resolution offers advantages over end-point measurements in plant quantitative genetics experiments, because growth dynamics can be assessed and analysed for genotype-phenotype association. Recently, network-based camera systems have been deployed as customizable, low-cost phenotyping solutions. Here, we implemented a large, automated image-capture system based on distributed computing using 180 networked Raspberry Pi units that could simultaneously monitor 1,800 white clover ( Trifolium repens ) plants. The camera system proved stable with an average uptime of 96% across all 180 cameras. For analysis of the captured images, we developed the Greenotyper image analysis pipeline. It detected the location of the plants with a bounding box accuracy of 97.98%, and the U-net-based plant segmentation had an intersection over union accuracy of 0.84 and a pixel accuracy of 0.95. We used Greenotyper to analyze a total of 355,027 images, which required 24–36 h. Automated phenotyping using a large number of static cameras and plants thus proved a cost-effective alternative to systems relying on conveyor belts or mobile cameras.

show abstract

Section: Introductionmentioning

confidence: 99%

Greenotyper: Image-Based Plant Phenotyping Using Distributed Computing and Deep Learning

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Al-Shakarji et al proposed an unsupervised segmentation approach that uses expectation maximization (EM) algorithm along with the K-means to get a quick result by performing calculations in a combined model to separate the front portion of an image from its background region [6]. When compared to K-means method, there will be less chance to get stop in a local optimum point by the use of EM algorithm.…”

Section: Unsupervised Segmentation Methodsmentioning

confidence: 99%

“…When compared to K-means method, there will be less chance to get stop in a local optimum point by the use of EM algorithm. In EM method, rather than assigning data point to only one cluster, we can assign it to different clusters in a partial manner [6]. For partial assignment, probabilistic distribution is used to model each cluster, so that each data point is assigned by a certain probability and finally a group of data points which is known as a cluster with the highest probability score will be selected.…”

Section: Unsupervised Segmentation Methodsmentioning

confidence: 99%

“…For partial assignment, probabilistic distribution is used to model each cluster, so that each data point is assigned by a certain probability and finally a group of data points which is known as a cluster with the highest probability score will be selected. In K-means method, the basic approach is to cluster similar points into separate clusters based on centroid and number of clusters [6].…”

Section: Unsupervised Segmentation Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A Performance Comparison of Supervised and Unsupervised Image Segmentation Methods

Baby

Devaraj

Mathew³

et al. 2020

SN COMPUT. SCI.

View full text Add to dashboard Cite

Image processing plays a vital role in many recent computer applications in the association with machine learning technology. The supervised training on dataset of features can only be successful if the segmentation process is accurate in the computer vision phase. The term segmentation is the process of extracting or identification of distinguishable regions in an image. This is performed based on the properties of image pixel intensity values and their proximities. This paper mainly focuses on an investigation of various latest image segmentation techniques performed in the field of computer vision and image processing. Segmentation plays a vital role in computer vision since any fault in segmentation will led to inaccurate extraction of features which results in wrong prediction of the decision support systems.

show abstract

“…In the field of segmentation [25–36], finely‐grained objects contain tenuous, flexuous, cross or pierced parts. Thus, for the sake of preserving the object's structure, one kind of strategy is to add the prior knowledge and constraints of the object's topology structure to the model.…”

Section: Related Workmentioning

confidence: 99%

Robust random walk for leaf segmentation

Chen

Zhang

et al. 2020

IET image process

View full text Add to dashboard Cite

In this study, the authors focus on the task of leaf segmentation under different imaging conditions (e.g. backgrounds and shadows). A new method ‐ robust random walk (RW) is proposed to propagate the prior of user's specified pixels. Specifically, they first employ RWs to take the relationship of pairwise pixels into consideration. A superpixel‐consistent constraint is added to make the edges of segmentation smooth. Owing to the effect of illumination, some parts of a leaf surface are brighter than others and it may further harm the subsequent label propagation. To address this problem, they learn a common subspace by taking into account the illumination of local and non‐local pixels. By doing so, it has good adaptability to process noise interfering and non‐uniform illumination. In addition, since RW only considers the pairwise relationship of pixels, it will be sensitive to the specified and connected pixels. Thus, they further employ a log‐likelihood ratio to predict the probability of a pixel belonging to the background and use it to guide the label propagation. Based on the proposed method, they can obtain a smoothed and robust leaf segmentation. Experimental results on unconstrained leaf images demonstrate the efficiency of their algorithm.

show abstract

Unsupervised Learning Method for Plant and Leaf Segmentation

Cited by 18 publications

References 15 publications

Greenotyper: Image-Based Plant Phenotyping Using Distributed Computing and Deep Learning

Greenotyper: Image-Based Plant Phenotyping Using Distributed Computing and Deep Learning

A Performance Comparison of Supervised and Unsupervised Image Segmentation Methods

Robust random walk for leaf segmentation

Contact Info

Product

Resources

About