Development of Paddy Rice Seed Classification Process using Machine Learning Techniques for Automatic Grading Machine

Table 3

Deep learning experimental results with VGG19 network locking 15 layers while training top layer set at 1024 nodes and varied input image size.

Model	Input image size	Acc200 (%)	Accl (%)	Test time per image (ms)	Model training time (hour)
VGG19	250	88.74	90.94	13.73	34 : 12
VGG19	200	86.76	90.95	9.61	25 : 13
VGG19	150	89.24	87.34	5.91	18 : 06
VGG19	100	84.41	83.20	5.30	13 : 26
VGG19	50	72.45	72.50	5.05	10 : 47