An Obstacle Recognizing Mechanism for Autonomous Underwater Vehicles Powered by Fuzzy Domain Ontology and Support Vector Machine

Table 3

The performance result of SVMs, BP algorithm, and the proposed scheme.

Obstacles classifier (object sample)	BP training time (s)	SVM training time (s)	Proposed system training time	BP testing time (s)	SVM testing time (s)	Proposed system testing time	BP accuracy	SVM accuracy	Proposed system accuracy
5–24	72.86	15.29	10.43	0.28	0.19	0.09	86.17%	90.87%	97.91%
10–24	153.76	19.93	9.47	3.91	2.28	0.14	85.12%	86.73%	98.17%
20–24	217.91	67.35	33.52	12.83	6.12	1.26	84.94%	89.90%	95.77%
5–36	141.26	47.82	26.51	2.93	1.06	0.07	87.95%	91.78%	98.31%
10–36	248.26	78.11	50.09	11.96	7.18	2.33	86.21%	92.99%	97.47%