Classification of Woven Fabric Faulty Images Using Convolution Neural Network

Table 3

Classification performance of the comparative methods.

Schemes	Precision	Recall	F1 measure	Accuracy
Hog-based KNN [36]	74.61	74.10	74.12	74.10
Walwet-based BPNN [37]	86.72	86.00	85.98	81.97
Kumar et al. [2]	79.3	79.1	80.2	79.7
Hu et al. [38]	87.4	87.9	83.5	85.7
Mak et al. [39]	82.6	78.0	83.5	80.8
Hu et al. [40]	75.5	71.4	87.9	79.7
Purposed CNN	83.66	83.5	83.56	94.46