Qualidade de Sementes

13 views
Skip to first unread message

Maria de Oliveira

unread,
Mar 26, 2020, 6:57:03 PM3/26/20
to FIELDimageR: A R package to analyze orthomosaic images from field trials.
Olá Filipe, 

Como havíamos conversado anteriormente, estou tentando quantificar a presença de Cerscospera nas sementes de soja e gostaria de saber se com o FieldimageR seria possível mensurar a intensidade dessa doença mesmo em imagens com o plano de fundo escuro como a em anexo? 
Você saberia me informar se tem alguma referência na literatura sobre o uso de FieldimageR para qualidade de sementes? 
Muito obrigada pela atenção.

Maria 
4.jpg

fmatias

unread,
Mar 27, 2020, 9:01:55 AM3/27/20
to FIELD...@googlegroups.com



Hi Maria, thank you for asking.

 

Please, see if the script below can help you.

 

Best,

Filipe


### Script – Seeds disease ###

# Packages:

library(FIELDimageR)

library(raster)

 

# Input image:

EX1<-stack("4.jpg")

plotRGB(EX1, r = 1, g = 2, b = 3)

 

# Choosing the index to remove the background:

EX1.I1<- fieldIndex(mosaic = EX1,index = c("BI","SCI","BGI","HUE"))

plot(EX1.I1$SCI)

 

# Removing the background (cropping above 0.4):

EX1.R1<- fieldMask(mosaic = EX1,index = "SCI",

                 cropValue = 0.4,

                 cropAbove = T)

 

# Seed selection:

par(mfrow=c(1,2))

plotRGB(EX1, r = 1, g = 2, b = 3)

plotRGB(EX1.R1$newMosaic, r = 1, g = 2, b = 3)

par(mfrow=c(1,1))

 

# Percentage of the image with seeds (22% of the image has seeds. This information can be used as a reference to compare images):

EX1.A1<-fieldArea(mosaic = EX1.R1$mask)

 

# Choosing the index to identify the disease area:

EX1.I2<- fieldIndex(mosaic = EX1.R1$newMosaic,index = c("BI","GLI","SI","VARI"))

 

# Affected region:

par(mfrow=c(1,2))

plotRGB(EX1.R1$newMosaic, r = 1, g = 2, b = 3)

plot(EX1.I2$BI)

par(mfrow=c(1,1))

 

# Identifying only disease area (crop below 0.7)

EX1.R2<- fieldMask(mosaic = EX1.R1$newMosaic,

                 index = "BI",

                 cropValue = 140,

                 cropAbove = T)

 

# Percentage area affected by the disease (27% of the seed area has dark regions. This information refers only to the proportion with seeds):

EX1.A2<-fieldArea(mosaic = EX1.R2$mask)

plotRGB(EX1.R2$newMosaic)

 

# An extra information about how viewing yellowed regions:

par(mfrow=c(1,2))

plotRGB(EX1.R1$newMosaic, r = 1, g = 2, b = 3)

plot(EX1.I2$SI)

par(mfrow=c(1,1))


 
Example of output:


LinkDin.jpg



Maria de Oliveira

unread,
Mar 27, 2020, 8:07:26 PM3/27/20
to FIELDimageR: A R package to analyze orthomosaic images from field trials.
Muito obrigada Filipe! Isto e exatamente o que eu estava procurando.
Reply all
Reply to author
Forward
0 new messages