A Composite Model of Wound Segmentation Based on Traditional Methods and Deep Neural Networks

Table 4

Comparison of existing methods. The numbers in parentheses are the numbers of images in the test sets.

Related work	Number of images	Approaches	Degree of complexity of the final solution	Precision(%)	mIoU(%)
Yadav et al. [3]	77(77)	Fuzzy c-means	simple	86.78	-
Dhane et al. [12]	105(105)	spectral clustering	simple	91.80	-
Dhane et al. [9]	70(70)	fuzzy spectral clustering	simple	87.30	79.0
Wang et al. [5]	650(150)	deep CNN	middle	-	47.30
Wang et al. [5]	950(190)	deep CNN	middle	93.43	73.36
Ours	950(190)	deep CNN+	complex	94.6906	85.8823