Content Comparison

Processed full-sized images

197,868

16.45 TiB

.jpgs only

Processed cutouts

1,472,436

1.56 TiB

Cropped segments of full-sized images. Includes jpgs, pngs, masks, and metadata json.

Cutouts available for hackathon

702,696

848.59 GiB

Cutouts available for hackathon. Includes jpgs, pngs, masks, and metadata json.

season

string

The growing season during which the cutout was captured (e.g., spring, summer). More Info

datetime

string (date-time)

The date and time when the cutout was captured in ISO 8601 format.

bbot_version

string

The version of the batch bot system used for processing the cutout. More info

batch_id

string

A unique identifier made up of a state abbreviation and date for the batch that contains the image (e.g., MD_2024-01-12).

image_id

string

A unique identifier made up of a state abbreviation and unix epoch timestamp for each image.

cutout_id

string

A unique identifier derived from the image_id and appended with an integer denoting the cutout number.

cutout_num

integer

The index or sequence number of the cutout within the image.

cutout_height

integer

The height of the cutout in pixels.

cutout_width

integer

The width of the cutout in pixels.

lens_model

string

The model of the camera lens used to capture the cutout.

validated

boolean

Indicates whether the cutout metadata has been validated.

cutout_version

string

data version number. Includes metadata and masks

cutout_props

object

Properties related to the cutout’s visual and structural features.

category

object

Category details for the cutout, including species classification.

is_primary

boolean

extends_border

boolean

bbox_area_cm2

number

blur_effect

number

num_components

integer

cropout_rgb_mean

array of 3 numbers

cropout_rgb_std

array of 3 numbers

non_target_weed

boolean

non target weed classifier prediction results

non_target_weed_pred_conf

number

prediction confidence of model indicating whether it’s a non-target weed. From custom trained classifier