Deep learning for detecting herbicide weed control spectrum in turfgrass

Table 1 The number of sub-images used to constitute the training, validation, and testing datasets of the herbicide weed control spectrum neural networks

Dataset	Weeds susceptible to ACCase-inhibiting herbicides				Weeds susceptible to synthetic auxin herbicides				No herbicide
Dataset	Crabgrass	Dallisgrass	Goosegrass	Tropical signalgrass	Dollarweed	Old world diamond-flower	Virginia buttonweed	White clover	Bermudagrass
Training	3000	3000	3000	3000	3000	3000	3000	3000	12,000
Validation	600	600	600	600	600	600	600	600	2400
Testing	600	600	600	600	600	600	600	600	2400

The herbicide weed control spectrum neural networks were trained to detect and discriminate the sub-images containing weeds susceptible to ACCase-inhibiting herbicides, weeds susceptible to synthetic auxin herbicides, or bermudagrass turf exclusively (no herbicide)

ISSN: 1746-4811