Table 1.

Rough classification of the tested approaches.

	GMM	Bayes	SPL	Cheddad Chen	SA1 SA2 SA3	DYC	SegNet U-Net DeepLab HardNet	PVT HSN
Preprocessing steps
None	x	x	x	x			x	x
Dynamic adaptation					x	x
Color space
Basic color spaces	x	x	x				x	x
Perceptual color spaces					x
Orthogonal color spaces						x
Other (e.g., color ratio)				x
Problem formulation
Segmentation based					x		x	x
Pixel based	x	x	x	x		x
Type of pixel classification
Rule based				x		x
Machine learning: parametric	x	x
Machine learning: non-parametric			x
Type of classifier
Statistical		x	x
Mixture techniques	x
Adaptive methods					x
CNN							x
Transformer								x