Leukocyte Segmentation Method Based on Adaptive Retinex Correction and U-Net

Table 2

Comparison of parameters of different methods under different datasets.

Datasets	Method	Dice	MIoU (%)	MPA (%)	Accuracy (%)
Cellavision datasets	DeepLabV3+	0.837	75.12	86.81	90.90
	PSPNet	0.828	74.69	86.06	89.49
	U-net	0.898	87.52	93.02	94.98
BCCD datasets	DeepLabV3+	0.804	76.48	85.19	88.33
	PSPNet	0.772	73.89	83.84	86.15
	U-net	0.901	88.75	92.55	94.27