RGB UAV plants imge dataset

Dataset Overview: Plant Species UAV Image Dataset (PSUID)

The Plant Species UAV Image Dataset (PSUID) is a comprehensive collection of high-resolution RGB images captured by an unmanned aerial vehicle (UAV) to facilitate the accurate identification and classification of plant species in heterogeneous plant areas. The dataset was meticulously collected and processed as part of a study conducted at Parco delle Cave in Brescia, Italy.

Dataset Highlights:

• Title: Plant Species UAV Image Dataset (PSUID)
• Focus Area: Parco delle Cave, Brescia, Italy
• Data Source: UAV-collected RGB images
• Classes and Descriptions:
  1. Agropyron repens
  2. Ailanthus altissima
  3. Arrhenatherum elatius 
  4. Artemisia verlotiorum
  5. Populus nigra
  6. Rubus caesius
  7. Ulmus minor

Methodology:

1. Study Area:

   • The study focuses on a one-hectare agricultural property within Parco delle Cave in Brescia, Lombardy, Italy.

2. Data Acquisition:

   • UAV flight on May 30, 2022, using a Mavic 2 Pro equipped with a Hasselblad L1D-20MP camera.
   • Flight parameters: 30 meters above ground, 4 ground control points, 319 images captured.

3. Image Processing:

   • Orthomosaic and digital surface model (DSM) generated using Agisoft Metashape.
   • Segmentation using eCognition software with multi-resolution and spatial difference segmentation.

4. Supervised Learning:

   • Training samples selected using color indicators for accurate species identification.
   • Classification using the K-nearest neighbor (KNN) algorithm in eCognition software.

5. Data Extraction and Preparation:

   • Plant species images extracted using ArcGIS software.
   • Creation of a comprehensive dataset with 782 images across seven plant species classes.

The data is published at https://www.mdpi.com/2504-446X/7/10/599