Descrição:
Thermographic Imaging of Physiological Transitions in Mice During Incremental Exercise / Mice Run Thermography Dataset:
This repository provides a complete data package related to mice running thermography. The package consists of four main items: a PDF description of the dataset, a CSV file containing the image metadata (REDU converts automatically to a tabular format), three random sample images, and the main ZIP file containing all the images.
Dataset Content and Origin --
This repository contains a ZIP file holding the entire dataset and an accompanying CSV file detailing its metadata. The data consists of 31,818 thermographic images captured during incremental treadmill running tests.
The tests were conducted on 11 "C57BL/6J" mice (120 days old) and were performed without electrical stimulus. Each recording was captured during an incremental running session where animals exercised until exhaustion.
The data were collected at the Laboratory of Applied Sports Physiology (University of Campinas — Unicamp), following ethical approval protocol CEUA 4940-1/2018.
Technical Specifications --
The dataset is provided in a single compressed file. The total of 31,818 entries are in JPEG format. Each file has a resolution of 360 x 306 pixels and an average size of approximately 30 KB.
Experimental Setup --
- Species: *Mus musculus* (C57BL/6J, 120 days old)
- Camera: FLIR One Pro (7.5–13 µm spectral range, 640×512 px)
- Position: 28 cm above the treadmill
- Treadmill stall dimensions: 9.1 cm width × 8 cm height × 18 cm length
- Recording rate: 30 Hz
- Location: Laboratory of Applied Sports Physiology — Unicamp
Proposed Applications --
This dataset is designed to support the classification and modeling of physiological transitions between intensity domains through thermographic imagery, while also enabling a broad range of research and development applications — such as thermal image classification and segmentation, exercise physiology modeling, domain transition detection, and data augmentation with cross-species learning — without being limited to these examples.