Open AccessLEGION BASED WEED EXTRACTION FROM UAV IMAGERY
Author(s) -
Saleh S. Alhewairini
Publication year - 2019
Publication title -
pakistan journal of agricultural sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.22
H-Index - 20
eISSN - 2076-0906
pISSN - 0552-9034
DOI - 10.21162/pakjas/19.8053
Subject(s) - multispectral image , computer science , artificial intelligence , computer vision , rgb color model , segmentation , pattern recognition (psychology) , remote sensing , geography
Multispectral images are the types of images which consist of many different bands of data for further processing. Vegetation Indexing has Different bands which are combined in the multispectral images and are used to accentuate the vegetated areas. In this one combination is most commonly used LIKE Ratio Vegetation Indexing (RVI). But using UAV imagery this problem has been solved because in UAV the area covered or captured by the cameras would be very wide so that the whole plot or crop can be viewed or captured from the UAV imagery. Legion is a Model used to do Scene Analysis Task, Image Segmentation, Object Selection and Speech Segregation. In this Research work we used the Legion Network approach to Detect or Segregate the Unwanted Crops or Weed to increase the crop productivity by Detecting or Segregating the Unwanted Crops or Weed. We used UAVDJI Phantom 4 having RGB Sensor and Tetracam NIR Sensor to collect the data set. Then we used ImageJ and MATLAB for applying Legion Model algorithm to process and analyze the multidimensional images using linear differential equations which solves the correlation problem of synchronization and Segregation of UAV based crop images to Detect the Unwanted crops