Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

An RGB orthomosaic of the study are, Codo in Catalonia (Spain).

Base map source: USGS National Map Viewer.

More »

Fig 1 Expand

Fig 2.

A canopy height model (CHM) derived from UAS-derived 3-D point cloud and a digital terrain model (DTM).

(A) UAS-derived 3-D point cloud. (B) Digital terrain model. (C) Canopy height model.

More »

Fig 2 Expand

Fig 3.

Object-based image analysis.

(A) RGB orthomosaic image with crown surfaces delineated by ITDe. (B) Image classification with NDVI for non-defoliated (in green) surfaces and defoliated / non-vegetation background (in black) with ITDe. (C) Colour-infrared composite using green, red and nir reflectance bands with crown surfaces delineated by ITDe. (D) Unsupervised classification with ExG for isolating shaded pixels (in grey). (E) Manual digitization of tree crowns (MCDe) for validating the results by ITDe. (F) Image classification with NDVI for non-defoliated (in green) surfaces and defoliated / non-vegetation background (in black) with MCDe.

More »

Fig 3 Expand

Fig 4.

Tree species identification and pine processionary moth defoliation in pines in the Codo forest site.

(A) Individutal automatic tree delination on the colour-infrared composite orthomosaic using green, red and nir reflectance bands. (B) Individutal automatic tree delination on the RGB orthomosaic. (C) Automatic tree classification in the field as holm oak or non-defoliated, partially defoliated and completely defoliated pine trhough multispectral high-resolution imagery.

More »

Fig 4 Expand

Table 1.

Confusion matrix for non-defoliated, partially defoliated and completely defoliated trees measured in the field or classified by using the multispectral imagery with the automatic ITDe.

More »

Table 1 Expand

Table 2.

Confusion matrix for non-defoliated, partially defoliated and completely defoliated trees measured in the field or classified by using the multispectral imagery with the manual onscreen-digitized tree segmentation (MCDe).

More »

Table 2 Expand

Fig 5.

Mean NDVI at the tree level.

(A) Mean NDVI values and standard error for pines (P) and holm oaks (Q). (B) completely defoliated (CD), partially defoliated (D) and non-defoliated pines (H) and holm oaks (Q) of the 110 randomly selected trees.

More »

Fig 5 Expand

Table 3.

Confusion matrix for classification assessment at the pixel level.

More »

Table 3 Expand

Fig 6.

Linear regression model between PPM defoliation measured in the field and through UAS multispectral images in Codo forest site with ITDe (A) and MCDe (B).

Green: non-defoliated trees; orange: partially defoliated; red: completely defoliated.

More »

Fig 6 Expand