Skip to main content

Table 1 Detailed Comparison of CotLeaf-1, CotLeaf-2 and CotLeaf-X Datasets. These three datasets differ in terms of season (Year, Y), growing environment (GH: Glasshouse, FD: Field), leaf number (L3 or L4), planting location, number of genotypes (see Table 2) or populations (A or B) imaged, total number of images, image location (see Fig. 1), presence/absence of multiple images per leaf, hairiness scale, and how hairiness scores were attributed

From: HairNet2: deep learning to quantify cotton leaf hairiness, a complex genetic and environmental trait

Characteristics

CotLeaf-1

CotLeaf-2

CotLeaf-X

[8, 22]

(this study)

(this study)

Year (Y)

Y1, Y2

Y3, Y4

Y3

Environment

GH, FD

GH (Y3 only), FD

FD

Leaf number

3, 4

3

3

Location

Narrabri, Canberra

Narrabri

Narrabri

Num. of Genotypes (G) or Populations (P)

27 (G)

27 (G)

2 (P)

Num. of Images

13597

810

5049 (A:3276, B:1773)

Image Loc.

First, Middle, Last

First

First, Blade

Multiple images/leaf

Yes

No

Yes (Pop. A)

No (Pop. B)

Genotype hairiness

Scale (GHS)

{’1’, ’2’, ’3’, ’3/4’, ’4’,

’4/4+’, ’4+’, ’5’, ’5+’}

{’1’, ’2’, ’3’, ’3/4’, ’4’,

’4/4+’, ’4+’, ’5’, ’5+’}

(Pop. A)

{’0’, ’1’, ’2’, ’3’, ’4’, ’5’},

(Pop. B)

{’2’, ’3’, ’4’, ’5’, ’5.5’}

Score attributed by

Genotype

Genotype

Individual image