FBK-TeV/WildBe

Wild Berry image dataset collected in Finnish forests and peatlands using drones

New V2

WildBe has Evolved!

Introducing WildBe-v2, the newly expanded dataset.

We're excited to announce a major update. The dataset has grown, offering more data and new possibilities.

Explore WildBe-v2 on Hugging Face

Introduction

Berry picking has long-standing traditions in Finland, yet it is challenging and can potentially be dangerous. The integration of drones equipped with advanced imaging techniques represents a transformative leap forward, optimising harvests and promising sustainable practices. We propose WildBe, the first image dataset of wild berries captured in peatlands and under the canopy of Finnish forests using drones. Unlike previous and related datasets, WildBe includes new varieties of berries, such as bilberries, cloudberries, lingonberries, and crowberries, captured under severe light variations and in cluttered environments.

How to use: an example of visualization

import json

import numpy as np
from datasets import load_dataset
from PIL import Image, ImageDraw

# Color map for classes
classes_color_map = {
    0: (225,15,10),
    1: (40, 150, 210),
    2: (10,0,210) ,
    3: (130,5,125) ,
}

# Load the dataset
dataset = load_dataset("FBK-TeV/WildBe", split="validation")

image = dataset[50]["image"]
labels = json.loads(dataset[50]["labels"])

draw = ImageDraw.Draw(image)

for label in labels:
    center_x = label["x"] * dataset[50]["width"]
    center_y = label["y"] * dataset[50]["height"]
    width = label["width"] * dataset[50]["width"]
    height = label["height"] * dataset[50]["height"]
    draw.rectangle(
        [
            (center_x - width / 2, center_y - height / 2),
            (center_x + width / 2, center_y + height / 2),
        ],
        outline=classes_color_map[label["class"]],
        width=2,
    )

image.show()

Data Fields

index: An integer representing the unique identifier for each example.
image: A PIL image.
split: A string indicating the data split, e.g., 'train', 'validation', or 'test'.
labels: A list of dictionaries, each containing:
  class: An integer representing the class identifier.
  label: A string representing the class name.
  x: A float representing the normalized x-coordinate of the center of the bounding box.
  y: A float representing the normalized y-coordinate of the center of the bounding box.
  width: A float representing the normalized width of the bounding box.
  height: A float representing the normalized height of the bounding box.
altitude: A float representing the altitude at which the image was taken (if available).
aperture: A float representing the aperture setting of the camera (if available).
area: A float representing the code of the geo-area in which the image was taken (if available).
date: A string representing the date when the image was taken (if available).
device: A string representing the device used to capture the image (if available).
exposure: A float representing the exposure time of the camera (if available).
focal: A float representing the focal length of the camera lens (if available).
height: An integer representing the height of the image in pixels.
width: An integer representing the width of the image in pixels.
iso: A float representing the ISO setting of the camera (if available).
latitude_deg: A float representing the latitude degree where the image was taken (if available).
latitude_dir: A string representing the latitude direction (if available).
longitude_deg: A float representing the longitude degree where the image was taken (if available).
longitude_dir: A string representing the longitude direction (if available).
source_image_id: A string representing the unique identifier for the source image from which the image was cropped.
time: A string representing the time when the image was taken (if available).

ArXiv link

https://arxiv.org/abs/2405.07550

APA Citaion

Riz, L., Povoli, S., Caraffa, A., Boscaini, D., Mekhalfi, M. L., Chippendale, P., Turtiainen, M., Partanen, B., Ballester, L. S., Noguera, F. B., & Poiesi, F. (2025). Wild Berry Image Dataset Collected in Finnish Forests and Peatlands Using Drones. In Computer Vision – ECCV 2024 Workshops. Springer Nature Switzerland.

Bibtex

@InProceedings{riz2024wild,
  title={Wild Berry Image Dataset Collected in Finnish Forests and Peatlands Using Drones},
  author={Riz, Luigi and Povoli, Sergio and Caraffa, Andrea and Boscaini, Davide and Mekhalfi, Mohamed Lamine and Chippendale, Paul and Turtiainen, Marjut and Partanen, Birgitta and Ballester, Laura Smith and Noguera, Francisco Blanes and others},
  booktitle="Computer Vision -- ECCV 2024 Workshops",
  year="2025",
  publisher="Springer Nature Switzerland",
}

Acknowledgement

The FEROX project has received funding from the European Union’s Horizon Framework Programme for Research and Innovation under the Grant Agreement no 101070440 - call HORIZON-CL4-2021-DIGITAL-EMERGING-01-10: AI, Data and Robotics at work (IA).

Partners