Image Semantic Recognition and Segmentation Algorithm of Colorimetric Sensor Array Based on Deep Convolutional Neural Network

Table 4

Comparative analysis for patterns.

Metrics	SCN	SOSD-Net	CBF	SFASM
Accuracy	0.7985	0.8201	0.8744	0.9194
Precision	0.8281	0.8614	0.9085	0.9415
F1-score	0.682	0.853	0.867	0.9195
Error ratio	18.17	15.63	10.15	7.129
Analysis time (s)	4.19	3.35	2.66	1.738

The proposed method achieves 8.84% high accuracy, 7.55% high precision, 11.88% high F1-score, 7.5% less error ratio and 8.15% less analysis time.