From 4c231c389d2f9e6e923a825d6baac5c7f8b34e79 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Miko=C5=82aj=20Data?= Date: Sat, 4 Nov 2023 22:09:46 +0100 Subject: [PATCH] PlantDieseaseDetection --- demos/PlantDieseaseDetection.ipynb | 47 ++++++++++++++++++++++++++++++ 1 file changed, 47 insertions(+) create mode 100644 demos/PlantDieseaseDetection.ipynb diff --git a/demos/PlantDieseaseDetection.ipynb b/demos/PlantDieseaseDetection.ipynb new file mode 100644 index 0000000..49bdb75 --- /dev/null +++ b/demos/PlantDieseaseDetection.ipynb @@ -0,0 +1,47 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "91a2e6f3-186e-4be5-9c83-96fc74f6892e", + "metadata": {}, + "source": [ + "# Plant disease\n", + "Sources:\n", + "- https://www.sciencedirect.com/science/article/abs/pii/S2352938521001361\n", + "- https://www.mdpi.com/2072-4292/14/23/5947\n", + "- https://link.springer.com/article/10.1007/s43154-020-00004-7/tables/1\n", + "\n", + "" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "4f42a100-a908-4538-b977-db81df1ce554", + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3 (ipykernel)", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.11.5" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +}

Disease

Autonomous

Material

Platform description

Spectral range

Data processing

References

Powdery mildew

Yes (green house)

Barley

Spectral imaging with automated sensor positioning system inside the greenhouse

VNIR (400–1000 nm)

Simplex volume maximisation (SiVM) and support vector machine (SVM)

[21••]

No (laboratory conditions)

Barley

Imaging setup with translation stage for sample presentation

VNIR (400–1000 nm)

Linear discriminant analysis (LDA) and feature selection with ReliefF

[25]

Grey mould leaf infection (Botrytis cinerea, fungi)

Semi (indoor)

Tomato

Plant growth chamber with additional lightening to ensure uniform illumination

5 bands: red, green, blue, near-infrared and red-edge

Self-organising classifier to classify healthy and infected tissue

[3]

Potato Y virus (Potyviridae, virus)

Semi (field condition)

Potato

A tractor-mountable measurement box carrying spectral imager, protection from external lighting and embedded PC

VNIR (400–1000 nm)

Deep learning (fully convolutional neural network)

[10•]

Tulip break virus

Semi (field condition)

Tulip

Field rail system with hand driven trolley platform

RGB combined with NIR

Deep learning (Faster R-convolutional neural network)

[9•]

Sclerotinia sclerotiorum (fungi)

No (laboratory conditions)

Oilseed rape

Indoor setup with translation stage used for imaging the plants

VNIR (384–1034 nm)

Partial least square discriminant analysis, SVM, radial basis function neural network, emerging learning neural network to detect disease

[23]

Apple scab (Venturia inaequalis, fungus)

No (laboratory conditions)

Apple

Indoor spectral imaging setup with translation stage for samples presentation

SWIR (1000–2500 nm)

Partial least square discriminant analysis (PLS-DA)

[22]

Anthracnose (Colletotrichum, fungi)

Yes (field)

Strawberry

A mobile (4 wheels) platform with mounted spectral sensor (non-imaging)

VNIR and SWIR (350–2500 nm)

Vegetation indexes, step wise discriminant analysis (SDA), Fisher discriminant analysis (FDA), k-nearest neighbours (kNN)

[20••]

No (laboratory conditions)

Strawberry

Imaging setup with translation stage for sample presentation

VNIR (400–1000 nm)

Spectral angle mapper (SAM), SDA, correlation measure (CM), partial least square regression (PLSR)

[26]

Downy mildew (Peronosporaceae, fungi)

Semi (green house)

Grapevine

Sensors and the light source arranged on a motorise line stage moving above the plants

Two systems: non-imaging: (350–2500 nm) and spectral imaging: (400–2500 nm) and (940–2550 nm)

SAM + 3 downy mildew indices

[27]

Early blight (Alternaria solani, fungi)

No (laboratory conditions)

Tomato

Imaging setup with translation stage for sample presentation

VNIR (380–1023 nm)

Extreme learning machine (ELM) classifier model, successive projections algorithm (SPA)

[24]

Fire blight (Erwinia amylovora, bacteria)

Semi (field condition)

Apple

Cameras mounted to an agricultural utility vehicle; an unmanned octocopter + multispectral camera

RGB combined with infrared + non-imaging VNIR and SWIR (350–2500 nm)

Vegetation indexes, PLSR and quadratic kernel support vector machine (QSVM)

[28]

Late blight (Phytophthora infestans, fungi)

No (laboratory conditions)

Tomato

Imaging setup with translation stage for sample presentation

VNIR (380–1023 nm)

Extreme learning machine (ELM) classifier model, successive projections algorithm (SPA)

[24]

Mosaic virus (various genera, virus)

No (laboratory condition)

Cucumber

Imaging setup with translation stage for sample presentation

946 nm to 2016 nm

PLS-DA, least square S-SVM

[29]

Target and bacteria spots

No (laboratory condition)

Tomato

Non-imaging spectrometer

350–2500 nm

Vegetation indexes

[30]

Cercospora leaf spot (Cercospora beticola)

No (laboratory condition)

Sugar beet

Imaging setup with translation stage for sample presentation

460–850 nm

Vegetation indexes and spherical k-means

[31]