The relationship between the plants and the environment is multitudinous and complex. They help in nourishing the atmosphere with diverse elements. Plants are also a substantial element in regulating carbon emission and climate change. But in the past, we have destroyed them without hesitation. For the reason that not only we have lost a number of species located in them, but also a severe result has also been encountered in the form of climate change. However, if we choose to give them time and space, plants have an astonishing ability to recover and re-cloth the earth with varied plant and species that we have, so recently, stormed. Therefore, a contribution has been made in this work towards the study of plant leaf for their identification, detection, disease diagnosis, etc. Twelve economically and environmentally beneficial plants named as Mango, Arjun, Alstonia Scholaris, Guava, Bael, Jamun, Jatropha, Pongamia Pinnata, Basil, Pomegranate, Lemon, and Chinar have been selected for this purpose. Leaf images of these plants in healthy and diseased condition have been acquired and alienated among two separate modules. Principally, the complete set of images have been classified among two classes i.e. healthy and diseased. First, the acquired images are classified and labeled conferring to the plants. The plants were named ranging from P0 to P11. Then the entire dataset has been divided among 22 subject categories ranging from 0000 to 0022. The classes labeled with 0000 to 0011 were marked as a healthy class and ranging from 0012 to 0022 were labeled diseased class. We have collected about 4503 images of which contains 2278 images of healthy leaf and 2225 images of the diseased leaf. All the leaf images were collected from the Shri Mata Vaishno Devi University, Katra. This process has been carried out form the month of March to May in the year 2019. The images are captured in a closed environment. This acquisition process was completely wi-fi enabled. All the images are captured using a Nikon D5300 camera inbuilt with performance timing for shooting JPEG in single shot mode (seconds/frame, max resolution) = 0.58 and for RAW+JPEG = 0.63. The images were in .jpg format captured with 18-55mm lens with sRGB color representation, 24-bit depth, 2 resolution unit, 1000-ISO, and no flash. Further, we hope that this study can be beneficial for researchers and academicians in developing methods for plant identification, plant classification, plant growth monitoring, leave disease diagnosis, etc. Finally, the anticipated impression is towards a better understanding of the plants to be planted and their suitable management.

Overview

The PlantDoc dataset was originally published by researchers at the Indian Institute of Technology, and described in depth in their paper. One of the paper’s authors, Pratik Kayal, shared the object detection dataset available on GitHub.

PlantDoc is a dataset of 2,569 images across 13 plant species and 30 classes (diseased and healthy) for image classification and object detection. There are 8,851 labels. Read more about how the version available on Roboflow improves on the original version here.

And here's an example image:

https://i.imgur.com/fGlQ0kG.png" alt="Tomato Blight">

Fork this dataset (upper right hand corner) to receive the raw images, or (to save space) grab the 416x416 export.

Use Cases

As the researchers from IIT stated in their paper, “plant diseases alone cost the global economy around US$220 billion annually.” Training models to recognize plant diseases earlier dramatically increases yield potential.

The dataset also serves as a useful open dataset for benchmarks. The researchers trained both object detection models like MobileNet and Faster-RCNN and image classification models like VGG16, InceptionV3, and InceptionResnet V2.

The dataset is useful for advancing general agriculture computer vision tasks, whether that be health crop classification, plant disease classification, or plant disease objection.

Using this Dataset

This dataset follows Creative Commons 4.0 protocol. You may use it commercially without Liability, Trademark use, Patent use, or Warranty.

Provide the following citation for the original authors:

@misc{singh2019plantdoc,
  title={PlantDoc: A Dataset for Visual Plant Disease Detection},
  author={Davinder Singh and Naman Jain and Pranjali Jain and Pratik Kayal and Sudhakar Kumawat and Nipun Batra},
  year={2019},
  eprint={1911.10317},
  archivePrefix={arXiv},
  primaryClass={cs.CV}
}

About Roboflow

Roboflow makes managing, preprocessing, augmenting, and versioning datasets for computer vision seamless.

Developers reduce 50% of their code when using Roboflow's workflow, automate annotation quality assurance, save training time, and increase model reproducibility.

This dataset is recreated using offline augmentation from the original dataset. The original dataset can be found on this github repo. This dataset consists of about 87K rgb images of healthy and diseased crop leaves which is categorized into 38 different classes. The total dataset is divided into 80/20 ratio of training and validation set preserving the directory structure. A new directory containing 33 test images is created later for prediction purpose.

Dataset

This dataset was created by AdityaMalhotra1412

Contents

Recent advances in computer vision and machine learning, most notably deep convolutional neural networks (CNNs), are exploited to identify and localize various plant species in salt marsh images. Three different approaches are explored that provide estimations of abundance and spatial distribution at varying levels of granularity in terms of spatial resolution. In the coarsest-grained approach, CNNs are tasked with identifying which of six plant species are present/absent in large patches within the salt marsh images. CNNs with diverse topological properties and attention mechanisms are shown capable of providing accurate estimations with > 90% precision and recall in the case of the more abundant plant species whereas the performance of the CNNs is observed to decline in the case of less common plant species. Estimation of percent cover of each plant species is performed at a finer spatial resolution, where smaller image patches are extracted and the CNNs tasked with identifying the plant species or substrate at the center of the image patch. In an ecological setting, several image patches (~100) are extracted and classified using this approach to estimate the percent cover of the various plant species in the image. For the percent cover estimation task, the CNNs are observed to exhibit a performance profile similar to that for the presence/absence estimation task, but with an ~ 5–10% reduction in precision and recall. Finally, estimation of the spatial distribution of the various plant species is performed via semantic segmentation of the input images at the finest level of granularity in terms of spatial resolution. The Deeplab-V3 semantic segmentation architecture is observed to provide very accurate estimations for abundant plant species; however, a significant degradation in performance is observed in the case of less abundant plant species and, in extreme cases, rare plant classes are seen to be ignored entirely. Overall, a clear trade-off is observed between the CNN estimation quality and the spatial resolution of the underlying estimation thereby offering guidance for ecological applications of CNN-based approaches to automated plant identification and localization in salt marsh images.

Indian Medicinal plant datasets is a repository that consists of medicinal plants images. The images are captured with varying background without any environment constraints

Fruit and vegetable plants are vulnerable to diseases that can negatively affect crop yield, causing planters to incur significant losses. These diseases can affect the plants at various stages of growth. Planters must be on constant watch to prevent them early, or infestation can spread and become severe and irrecoverable. There are many types of pest infestations of fruits and vegetables, and identifying them manually for appropriate preventive measures is difficult and time-consuming.This pretrained model can be deployed to identify plant diseases efficiently for carrying out suitable pest control. The training data for the model primarily includes images of leaves of diseased and healthy fruit and vegetable plants. It can classify the multiple categories of plant infestation or healthy plants from the images of the leaves.Licensing requirementsArcGIS Desktop — ArcGIS Image Analyst extension for ArcGIS ProArcGIS Enterprise — ArcGIS Image Server with raster analytics configuredArcGIS Online — ArcGIS Image for ArcGIS OnlineUsing the modelFollow the guide to use the model. Before using this model, ensure that the supported deep learning libraries are installed. For more details, check Deep Learning Libraries Installer for ArcGIS. Note: Deep leaning is computationally intensive, and a powerful GPU is recommended to process large datasets.Input8 bit, 3-band (RGB) image. Recommended image size is 224 x 224 pixels. Note: Input images should have grey or solid color background with one full leaf per image. OutputClassified image of the leaf with any of the plant disease, healthy leaf, or background classes as in the Plant Leaf Diseases dataset.Applicable geographiesThis model is expected to work well in all regions globally. However, results can vary for images that are statistically dissimilar to training data.Model architecture This model uses the ResNet50 model architecture implemented in ArcGIS API for Python.Accuracy metricsThis model has an overall accuracy of 97.88 percent. The confusion matrix below summarizes the performance of the model on the validation dataset. Sample resultsHere are a few results from the model:Ground truth: Apple_black_rot / Prediction: Apple_black_rotGround truth: Potato_early_blight / Prediction: Potato_early_bightGround truth: Raspberry_healthy / Prediction: Raspberry_healthyGround truth: Strawberry_leaf_scorch / Prediction: Strawberry_leaf_scorch

PlantDoc is a dataset for visual plant disease detection. The dataset contains 2,598 data points in total across 13 plant species and up to 17 classes of diseases, involving approximately 300 human hours of effort in annotating internet scraped images.

The dataset contains aerial agricultural images of a potato field with manual labels of healthy and stressed plant regions. The images were collected with a Parrot Sequoia multispectral camera carried by a 3DR Solo drone flying at an altitude of 3 meters. The dataset consists of RGB images with a resolution of 750×750 pixels, and spectral monochrome red, green, red-edge, and near-infrared images with a resolution of 416×416 pixels, and XML files with annotated bounding boxes of healthy and stressed potato crop.

Dataset of images of dragon fruit plants, collected from different media and taken from a dragon fruit field in Rio Branco, Brazil, with a total of 600 images classified among 300 photos of sick plants, with fish eyes among others and 300 photos of healthy plants. For many of the photos, a simple smartphone camera was used to capture the images.

There are two datasets and one table uploaded in this platform under the title "MED117_Medicinal Plant Leaf Dataset & Name Table". A folder is created with title "MED 117 Leaf Species". Inside this two sub folders with titles " Raw leaf image set of medicinal plants_v2" and "Segmented leaf set using UNET segmentation" are created. Raw leaf image set consists of leaf images of 117 medicinal plants found in Assam. All the samples are collected by visiting different (Govt, Public and Private) medicinal gardens situated in different places of Assam and some other general places where they are mostly found. Videos of 10 to 15 seconds duration were taken for two to three leaves of every species on a white background and video recording was done using a SLR Canon Camera. Individual videos were segregated into image frames and thus were able to get around 77,700 jpg image frames from the videos. The Raw leaf image set consists of folders with scientific name and common name within bracket. Second folder with title "Segmented leaf set using UNET segmentation" consists of 115 medicinal plant species with their segmented leaf image samples using UNET segmentation technique. Here two species are excluded from the original dataset due to small unpredictable size of the samples, so total 115 subfolders inside the segmented folder is achieved. Thirdly a table in doc format with title "Medicinal Plant Name Table" is uploaded and it includes Scientific name, Common name and Assamese name of the plants listed in the folders in the same sequence. The whole contribution is absolutely original and new, collected from different sources then processed for segmentation and prepared the table by discussing with taxonomy experts from Botany department of Gauhati University, Guwahati, Assam. India.

A dataset of 61,486 images of plant leaves and backgrounds, with each image labeled with the disease or pest that is present. The dataset was created by researchers at the University of Wisconsin-Madison and is used for research in machine learning and computer vision tasks such as plant disease detection and pest identification.

Basil/Tulsi Plant is harvested in India because of some spiritual facts behind this plant,this plant is used for essential oil and pharmaceutical purpose. There are two types of Basil plants cultivated in India as Krushna Tulsi/Black Tulsi and Ram Tulsi/Green Tulsi.Many of the investigator working on disease detection in Basil leaves where the following diseases occur 1) Gray Mold 2) Basal Root Rot, Damping Off 3) Fusarium Wilt and Crown Rot4) Leaf Spot5) Downy MildewThe Quality parameters (Healthy/Diseased) and also classification based on the texture and color of leaves. For the object detection purpose researcher using an algorithm like Yolo, TensorFlow, OpenCV, deep learning, CNNI had collected a dataset from the region Amravati, Pune, Nagpur Maharashtra state the format of the images is in .jpg.

Context

Humans in the Loop is happy to publish open access segmentation masks for a dataset provided by the Computer Vision and Biosystems Signal Processing Group at the Department of Electrical and Computer Engineering at Aarhus University.

Humans in the Loop has performed full semantic segmentation using masks on the images in 2 classes: plant and background (soil, container). The masks are ultra-precise and follow each plant’s leaf structure.

Content

The dataset contains 144 images of plant seedlings from 3 containers shot at different time intervals within the span of 2 months. Each container contains up to 40 single plants, each one of which has been marked with a bounding box for better visibility.

The images are segmented in 2 classes:

000000⬛- background

78C814🟩 – plant

Acknowledgements

The images were kindly annotated by the trainees of the Preemptive Love Coalition’s WorkWell program in Iraq.

Together with the Agriculture University, we compiled a database of plant images and omics data. The dataset contains images of four distinct plant maladies, including powdery mildew, rust, leaf spot, and blight, as well as gene expression and metabolite data. Using a high resolution camera in a controlled environment at the facility of the Agriculture University of Peshawar, we captured 8,000 images of plants, with 2,000 images for each disease type. Each image was labeled with the disease type corresponding to it. The images were preprocessed by resizing them to 224x224 pixels and standardizing the pixel values. The dataset was divided into 70:15:15 training, validation, and testing sets, correspondingly. In addition to collecting images of the same plants, we also collected gene expression and metabolite data. We extracted RNA from the plant leaves using a commercial reagent and sequenced it on an Illumina HiSeq 4000 platform. The average length of the 100 million paired-end readings obtained was 150 base pairs. The unprocessed reads were trimmed with Trimmomatic and aligned with STAR against the reference genome. We counted the number of reads that mapped to each gene using featureCounts, and then identified differentially expressed genes between healthy and diseased plants using the DESeq2 package in R. Using gas chromatography-mass spectrometry (GC-MS), we gathered additional metabolite information. Using a methanol-water extraction protocol, we extracted metabolites from the plant leaves and analyzed the extracts using GC-MS. We obtained 500 metabolite characteristics, including amino acids, organic acids, and sugars.If you use the dataset mentioned here, please make sure to give credit to the researchers by citing their paper titled 'Deep Learning for Plant Bioinformatics: An Explainable Gradient-Based Approach for Disease Detection.'ReferenceShoaib, M., Shah, B., Sayed, N., Ali, F., Ullah, R., & Hussain, I. (2023). Deep learning for plant bioinformatics: an explainable gradient-based approach for disease detection. Frontiers in Plant Science, 14(October), 1–17. https://doi.org/10.3389/fpls.2023.1283235

Images and data for the study "Plant image identification application demonstrates high accuracy in Northern Europe"

Details: Jaak Pärtel, Meelis Pärtel, Jana Wäldchen, Plant image identification application demonstrates high accuracy in Northern Europe, AoB PLANTS, Volume 13, Issue 4, August 2021, plab050, https://doi.org/10.1093/aobpla/plab050

The data table displays Flora Incognita's identification results together with species and observations characteristics. All (3199) used images are included.

The study was conducted in two parts: database and field study.

Database study images have been taken from eBiodiversity database (https://elurikkus.ee/en) under Creative Commons Attribution 4.0 International (CC BY 4.0) licence (https://creativecommons.org/licenses/by/4.0/). Please cite the original source for the images as well when using the dataset.

Field study images were taken by Jaak Pärtel in 2020 in field conditions from different habitats across Estonia.

Here we provide an updated image dataset and supporting data files, version 2, for the following primary article. Please refer to the primary article as well as the supporting data and updates provided here for all details.Wilf P, SL Wing, HW Meyer, J Rose, R Saha, T Serre, NR Cúneo, MP Donovan, DM Erwin, MA Gandolfo, E González-Akre, F Herrera, S Hu, A Iglesias, KR Johnson, TS Karim, X Zou. 2021. An image dataset of cleared, x-rayed, and fossil leaves vetted to plant family for human and machine learning. PhytoKeys 187: 93–128, doi:10.3897/phytokeys.187.72350The dataset version that corresponds exactly to the published article remains archived here as version 1 and is easily accessed by toggling the dataset version in this window.The total image-collection size is now 34,368, consisting of 30,252 images of cleared and x-ray leaves and 4,116 of fossils.Change list, version 1 to version 2:1) Addition of NMNS Cleared Leaf Database (4,076 images).The most significant change in version 2 is the addition of 4,076 images from the National Museum of Nature and Science (NMNS, Ibaraki, Japan) Cleared Leaf Database, made possible by the kind assistance of Dr. Atsushi Yabe, who is included here as a coauthor of version 2. The collection was made by Drs. Toshimasa Tanai of Hokkaido University and Kazuhiko Uemura of NMNS. More information on the NMNS Cleared Leaf Database and one-at-a-time image access are available at the website: https://www.kahaku.go.jp/research/db/geology-paleontology/cleared_leaf/index.php?lg=en . A prior publication using the database is Iwamasa and Noshita (2023), PLoS Comput Biol 19: e1010581.All taxonomic names for the NMNS specimens as given were vetted and updated to species level by Edward Spagnuolo (acknowledged for his kind assistance) and P. Wilf using the Taxonomic Names Resolution Service (TNRS) and other sources (we note that the names attached to the prior cleared and x-rayed leaf images carried over from version 1 remain vetted only to family level, although taxa of interest can be easily updated using TNRS and many other resources). The vetted names were then used to update the as-given NMNS filenames to the same alpha-sortable format as the prior images (Family_Genus_species_dataset_catalognumber.jpg) and integrated into the same family folders for maximum ease of use. We thank Ivan Rodríguez for his kind assistance with this step.2) Filename cleanup in all directories and updates to affected catalog files.Thousands of filenames and their associated catalog entries were improved by batch-removing all periods and spaces (i.e. "sp." and "sp. ", "cf.", "aff.", "x. ") and cross-checked for consistency.3) A small number of new fossils were added, namely 34 leaves from Dipterocarpaceae and other families from the Pliocene of Brunei (Wilf et al. 2022, PeerJ and online supplement) and seven leaves of Macaranga kirkjohnsonii from the Eocene Laguna del Hunco flora, Chubut, Argentina (Wilf et al. 2023, Am. J. Bot. and online supplement).4) The following nomenclatural updates were applied to the filenames of all affected images (fossils and extant) and related catalog entries:Adoxaceae to Viburnaceae.Vauquelinia coloradensis to Kageneckia coloradensis (after Denk et al. 2023).Vauquelinia lineara to Vauquelinia liniara (typo correction).Browniea, Camptotheca, and the five living Nyssaceae genera are all now categorized in Nyssaceae (some were in Cornaceae in v. 1).File annotationsThe version 2 files are provided here as zip archives, as follows. As noted above, the version 1 files remain available by toggling the database version.Extant_Leaves_A-E_v2.0.zipExtant_Leaves_F-O_v2.0.zipExtant_Leaves_P-Z_v2.0.zipFamilies A–E, F–O, and P–Z, respectively, of cleared and x-rayed leaf images (30,252 images).Florissant_Fossil_v2.0.zipFossil-leaf image collection from Florissant Fossil Beds National Monument (3,320 images).General_Fossil_v2.0.zipFossil-leaf image collection from several other sites (796 images).General_Fossil_uncropped_v2.0.zipReference set of most of the uncropped image versions for the General Fossil collection, for access to scale bars and other archival information not otherwise available digitally (see main article and supplements linked in item 3 above). Filenames are suffixed with "_uncropped" and may have minor differences in format from the cropped set.supplemental_data_v2.0.zipArchive containing three files:Master_inventory_leavesdb_v2.0Master inventory file listing all extant and fossil specimens.See details in the main article (esp. table 1) for how to look up additional specimen data, which are easily available on the Web for most of the collections using the catalog numbers listed in this inventory file (also see below). Please note that the catalog numbers listed here may be primary or secondary, as described in the main article (table 1). The "old_Family" field preserves legacy data that can assist in locating physical specimens in the collections, which usually retain their original taxonomic organization (see main text).The other two files are catalogs of specimen data not otherwise available on the Web (see main article).General_fossils_catalog_v2.0.csvSpecimen data for the "General fossil" image collection. As mentioned in the primary article, several fossils retain their generic names, even if they are known to be botanically incorrect in publications or in the opinion of the present authors and thus placed in scare quotes in the primary article. In this case, the listed family name is regarded as correct. Scare quotes cannot be used in filenames and are thus omitted.Wing_x-ray_catalog_v2.0.csvVoucher data for the Wing X-Ray image collection.Technical notes for the Wing x-rays:Catalog number field in the Master Inventory file = negative number + leaf number as listed in this file.Example: "Wing_199-001" in the Master Inventory = negative 199, leaf 1 here =Alphonsea arborea(Annonaceae) = primary voucher US 904529.Some typographical errors in this legacy catalog are left as-is, and identifications are not updated here. Vetted spellings and updated family and order assignments can be found by catalog number (= negative + leaf number) in the Master Inventory file. This file includes some additional records that did not meet criteria for the image dataset.

Image datasets play a crucial role across diverse fields, including computer vision, machine learning, medical research, and social sciences. These datasets serve as a valuable resource, providing rich visual information that enables researchers, developers, and professionals to train and validate their models, algorithms, and theories. In the agricultural domain, a specific image dataset focused on sugarcane leaf diseases holds significant importance. Such datasets offer researchers, agronomists, and farmers a valuable tool to identify, classify, and study various leaf diseases affecting sugarcane crops. By analyzing these images, experts can develop more accurate disease detection algorithms and early warning systems, facilitating prompt disease management and preventing widespread crop damage and yield loss. Additionally, a comprehensive dataset allows for the exploration of disease patterns, environmental factors, and potential mitigation strategies, thereby advancing research and improving overall crop management practices to ensure the health and productivity of sugarcane crops. The Sugarcane Leaf Dataset consists of 7134 high-resolution images of sugarcane leaves stored in JPEG format, with dimensions of 768 × 1024 pixels. Categorized into 12 distinct classes, including 10 disease categories, a healthy leaves category, and a dried leaves category, the dataset covers a wide range of common sugarcane leaf diseases, ensuring easy access and identification of specific disease samples. These images were collected through extensive field surveys, capturing different angles and stages of the diseases and guaranteeing a comprehensive representation of visual characteristics. With their high-quality resolution of 72 dots per inch (dpi), the images in the dataset provide clear and detailed visual representation of the sugarcane leaf samples.

This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 United States License. # Data origins The dataset is originally hosted at PlantVillage Disease Classification Challenge. We use the modified version in this github repository to do controlled experiments. We only use the raw color images dataset and delete the unconventional characters in the classes directory name and .csv filenames. # Directory explanation The 80-20 direcotry has multiple .txt files which contain the training (~80%), validation(~10%) and testing (~10%) datasets instances filenames and the corresponding label indexes. The validation dataset quantity is 5430 in all data separation. In our experiment code (not included in this archive), the validation and testing dataset are merged together. # Data usage ## Replicate our experiments We have used this dataset in writing our paper. The reference information can be seen at https://gitlab.com/huix/leaf-disease-plant-village. ### Steps 1. cd to the direcotry (e.g. /home/usrname/plantvillage_deeplearning_paper_dataset) that contains the color directory. 2. run python change_filename_prefix.py --prefix /home/usrname/plantvillage_deeplearning_paper_dataset to modify the prefix path (which is /home/h/plantvillage_deeplearning_paper_dataset in our former generated datasets). 3. Fin. You can use our opens ource codes repository to do the later experiments. ## Generate your own training/validation/testing datasets This data separation generating code isn't included in the dataset archive, it is in our open source code. Please see our open source code repository for the detailed information. If you have any questions, you can contact the author through email. The email address is a QR code in the archive. {"references": ["Hughes, D.P., Salathe, M.: An open access repository of images on plant health to enable the development of mobile disease diagnostics. ArXiv e-prints (2015). 1511.08060", "Mohanty, S.P., Hughes, D.P., Salath\u00e9, M.: Using deep learning for image-based plant disease detection. Frontiers in Plant Science 7, 1419 (2016). doi:10.3389/fpls.2016.01419"]} https://gitlab.com/huix/leaf-disease-plant-village