Research on Indoor Scene Classification Mechanism Based on Multiple Descriptors Fusion

Table 5

Indoor scene classification execution time.

	Step	Parameters	Time (s)
Descriptor generation	Extracting D-SIFT feature	imageSize = 640 ∗ 480	0.0840
	Mapping feature point	S = 20	0.0096
		S = 50	0.0140
		S = 100	0.0218
	Counting histograms	H = 0	0.0006
		H = 1,C = 1	0.0004
		H = 2, C = 1 or H = 1, C = 2	0.0003
		H = 2, C = 2	0.0002
Descriptor classification	Classifying the input descriptor		0.0010
			0.0016
			0.0029
			0.0062
			0.0131
			0.0291