Tree Image Dataset for Biophysical Parameter Estimation using Stereoscopic Vision

Published: 13 January 2023| Version 3 | DOI: 10.17632/nx3ggv7pxf.3
Cedric Kiplimo,
Billy Okal


This dataset contains 115 stereoscopic image pairs of trees within the Heroes Garden and surrounding parks at the Dedan Kimathi University of Technology, Nyeri, Kenya. In addition, 61 calibration image pairs have also been provided for the benefit of those who prefer to perform camera calibration on their own. The dataset is uploaded as a single Zip file, which upon extracting unfolds into five folders and three CSV files. All filenames for images in the ‘breast-height’ and ‘trunks’ folders follow the form TREE_ID-DISTANCE-PAIR_ID-NUMBER. DISTANCE refers to the distance between the tree and the camera, and is not included in the filenames for images in the 'full-trees' folder since these distances were not measured. 1. breast-height (folder): Contains 25 image pairs of tree trunks wrapped with white masking tape around the circumference at the breast height position (1.3 m above the trunk base). These images can be used to study the performance of a technique/algorithm based on stereoscopic photogrammetry at locating the breast height. 2. full-trees (folder): Contains 10 image pairs of full trees (shorter than 6m). Using these images, an algorithm based on stereoscopic photogrammetry may be developed to extract the heights and crown diameters of these trees. 3. Trunks (folder): Contains 80 image pairs of tree trunks belonging to 20 trees. They can be used to develop an algorithm based on stereoscopic photogrammetry to estimate the diameters at breast height (DBH) of these trees. 4. Calib-v1 (folder): Contains 30 calibration image pairs from which calibration parameters can be found. These parameters should be applied only to the images in the 'trunks' and 'full-trees' folders. If you prefer to avoid the trouble of camera calibration, this folder also contains the files 'stereo_params_v1.yml' and 'stereo_params_v1.txt' where the calibration parameters have been saved in a machine-readable and a human-readable format respectively. 5. Calib-v2 (folder): Similar to 'Calib-v1' except that the camera parameters apply to the images in the 'breast-height' folder only. It contains 31 calibration image pairs. 6. dbh_values.csv (file): Contains the ground truth values of the DBHs of the trees in the 'trunks' folder. 7. crown_diameter_values.csv (file): Contains the ground truth values of the crown diameters of the trees in the 'full-trees' folder. 8. tree_height_values.csv (file): Contains the ground truth values of the heights of the trees in the 'full-trees' folder. Before capturing the images in this dataset, camera calibration was performed using the well-known OpenCV library. For the meaning of the individual camera parameters in the text and YAML files, please refer to this article by OpenCV:


Steps to reproduce

1. Build a stereo camera 2. Write python code for capturing left and right image pairs (Use the OpenCV library 3. Capture the images


Dedan Kimathi University of Technology School of Engineering


Stereopsis, Forest Inventory, Computer Vision Algorithms