Advanced Data Analysis Techniques for Weed Detection in Agriculture

Desklib provides solved assignments and past papers to help students succeed.

Table 1: DWC taxonomy
Factors/Class Main Attributes Common Instances
Data
Image collection Crop images, Coloured image,
normalized image, Multi spectral
images
Digital camera, Datasets, HDR
camera, four μEye camera,
spectrograph, charge-coupled device
(CCD) camera, multiSPEC 4C,
Unmanned aerial vehicle (UAV)
camera
Image processing Weed mapping, Bitmap format,
size reduction, segmentation
PC computer, MATLAB software,
Agisoft PhotoScan, eCognition
Developer 8 software, Adaptive
segmentation algorithm, Particle
swarm optimum
Weed detection
Feature extraction Morphology, visual textures,
shape, spatial contexts
Fourier descriptors, moment invariant
features, Zernike moments,
Histogram, Hyperspectral imaging,
Fourier descriptors, Hough transform
algorithm (HT), AdaBoost algorithm,
Gray-Level Co-occurrence Matrix
(GLCM)
Fuzzification Fuzzy values, fuzzy inference
mechanism
Fuzzy logic Toolbox, Fuzzy logic
controller, Inference engine, binary
matrix, fuzzy logic real time system,
Raspberry Pi, Conv1D-based model,
k-means clustering
Classification
Weed classification Algorithms Machine learning algorithms, Shape
based weed detection (SBWD)
algorithm, back-propagation learning
algorithm, deep learning, Landsat
Enhanced Vegetation Index (EVI),
Simple Linear Iterative Clustering

(SLIC), Binary Classification
Confusion Matrix
Classifiers K-Nearest Neighbour, Support Vector
Machine, Artificial neural network,
AdaBoost Softmax classifier, Color
co-occurrence matrices (CCM),
Object Based Image Analysis(OBIA)
Weed control Weed coverage, distribution,
weed patch,
Robotic sprayer, machine vision
algorithms, herbicide application
system, robotic weed control system,
real time kinematic GPS (RTKGPS),
Arduino script