Descrição:
This repository contains a dataset of high-quality images of sugarcane stems from two varieties—Vertix 03 and Vertix 12—collected at approximately ten months of growth. The images were produced under controlled laboratory conditions following a dedicated acquisition protocol specifically developed to ensure high image quality and consistency.
The dataset was originally created for Lemos’ doctoral research, in which it served as the foundation for training and evaluating artificial intelligence models aimed at varietal classification of sugarcane. To the best of our knowledge, all samples were obtained from healthy plants.
Each image depicts a single sugarcane stem photographed against a neutral black background under standardized lighting conditions. The images are stored in JPEG format, with a resolution of approximately 24 megapixels (6000 × 4000 pixels) and an average file size of about 5 megabytes. No post-processing was applied.
The dataset is organized as follows:
• 1 PDF file (this document): Description and metadata of the repository
• 4 JPEG files: Random image samples illustrating the dataset contents
• 1 ZIP file: Images of the empty background
• 1 ZIP file: Images of the studio and the image acquisition setup
• 2 ZIP files: Images of Vertix 03 (parts 1 and 2), for focal distances of 35 mm
• 2 ZIP files: Images of Vertix 03 (parts 1 and 2), for focal distances of 50 mm
• 1 ZIP file: References for Vertix 03 *
• 2 ZIP files: Images of Vertix 12 (parts 1 and 2), for focal distances of 35 mm
• 2 ZIP files: Images of Vertix 12 (parts 1 and 2), for focal distances of 50 mm
• 1 ZIP file: References for Vertix 12 *
* Reference images, such as sugarcane stems photographed alongside rulers or pens for scale calibration.
This dataset is intended primarily for research in varietal classification and related fields such as computer vision, plant phenotyping, and machine learning applied to agricultural analysis. Details about the photographic setup and acquisition parameters are provided in the following sections of this document.