A Fast Decision Algorithm for VVC Intra-Coding Based on Texture Feature and Machine Learning

Table 2

Experimental results comparing the proposed algorithm with VTM 10.0.

Sequence type	Test sequence	BD-BR (%)	△T (%)
A1(3840 × 2160)	Tango2	1.23	54.25
	Food market4	0.78	55.63
	Campfire	1.18	52.22
A2 (3840 × 2160)	Cat robot1	1.34	53.8
	Daylight road2	1.51	53.62
	Park running3	1.23	51.32
B (1920 × 1080)	Basketball drive	2.83	60.67
	BQ terrace	1.76	55.37
	Cactus	1.57	50.63
	Kimono	1.08	50.18
	Park scene	0.97	49.23
C (832 × 480)	Basketball drill	1.94	52.43
	BQ mall	1.76	50.22
	Party scene	0.91	50.07
	Race horses C	1.08	49.23
D (416 × 240)	Basketball pass	1.43	55.24
	BQSquare	0.86	49.47
	Blowing bubbles	1.03	49.91
	Race horses	0.94	53.17
E (1280 × 720)	Four people	1.62	53.35
	Johnny	1.73	54.31
	Kristen and Sara	1.16	53.48
	Average	1.36	52.63