A geospatial Python class for crop recognition over drone orthophoto – Tutorial

Hydrate

Spatial analysis and machine learning sometimes require massive coding in order to achieve decent results such as identifying plants from a drone orthophoto. We wanted to create a simple workflow for beginner and intermediate Python users to work with these libraries without much pain or frustration. This tutorial has the complete procedure to use a Python class that recognizes plants from an orthophoto based on sample points and creates intermediate plots and identifies plants as point shapefiles.

Tutorial

Code

from workingTools import cropRecognition
import matplotlib.pyplot as plt

Create crop object

crop = cropRecognition()

Define raster and sample plants

crop.defineRaster('../odm_orthophoto/agaveCrop.tif')
CRS of Raster Data: EPSG:4326
Number of Raster Bands: 3
Interpretation of Raster Bands: (<ColorInterp.red: 3>, <ColorInterp.green: 4>, <ColorInterp.blue: 5>)
crop.definePoints('../Shp/agavePoints_3p.shp')
CRS of Point Data: epsg:4326

Retrieve information about the raster and plants

# get raster crs
crop.cropRaster.crs
CRS.from_epsg(4326)
# get point coordinates
crop.pointCoords
[(-104.60251209450945, 21.488914700478507),
 (-104.6026805707282, 21.488557889609346),
 (-104.60290182768277, 21.488567022907)]

Define parameter for the match template and raster band

# ratio of analysis
crop.pointRatio = 40
# raster band for the match template analysis
crop.selectedBand = crop.blueBand
# sample row, col and array for each point
crop.getPointRowCol()
Point N°:0 corresponds to row, col: 161, 6842
Point N°:1 corresponds to row, col: 3162, 5517
Point N°:2 corresponds to row, col: 3085, 3776
# plant arrays
crop.plotReferenceImages()

Perform single match template

# define object order and output raster path
matchTemplate2 = crop.singleMatchTemplate(2,'../Rst/matchTemplate2.tif')
# plot match template
#plt.imshow(matchTemplate0)

Match template analysis for all points

# define method (quantile or threshold) and value
crop.pointsMatchTemplate('threshold',0.32)
Processing image Nº 0
Processing image Nº 1
Processing image Nº 2
# plot all templates 
#crop.plotAllTemplates()

# save match template to a shapefile
crop.saveMatchShp('../Shp/matchCrop.shp')

Cluster analysis

# Birch clustering
crop.birchFilter()
# plot birch points
crop.plotBirchPoints()
# save cluster points as shapefile
crop.saveBirchShp('../Shp/birchCrop.shp')

Input file

You can dowload the input files from this link.

The code with sample orthophoto can be downloaded from this Github repository:

https://github.com/SaulMontoya/spatialPythonClassCropRecognitionDroneOrtophoto

—————
TRANSCOM ISPFree Sigma HSE Email
Level 6 Domain Names are FREE – Register Now.

Related Posts

Boost Inflight Internet- Business Hosting- Secure Email Account- Dropcatch Domain Names- Antisnoop Email- Free Secure Email- Cheap VOIP Calls- Free Hosting and Email- Aero Connectivity- Premium Domains for Sale- Transcom Telecom- Satphone Airtime- Mobile Plans- Free Domain Names- Organic Products- Aviation News
Transcom ISP - Transcom VOIP - Free Secure Email - Dropcatch Software - FastApn Inflight - Aero Connect - Premium Domains