Multi-Label Peach RGB-D Dataset

Multi-Label Peach RGB-D Dataset

(Including RGB Images, Depth Images, and Infrared Images)

Authors: Yuan Rao / Qing Luo / Peilin Huo / Yipu Li / Jingyao Zhang

The ML Peach RGB-D Dataset consists of 2,050 multimodal peach tree images captured using the Microsoft Azure Kinect DK. Each image includes 3 different modalities: RGB images, depth images, and infrared images.

All images are aligned with the RGB images and manually labeled into 4 categories: unoccluded, occluded by leaves, occluded by branches, and occluded by fruit.

Annotations in “.txt” format can be found in the “annotation_yolo” folder. Camera parameters for each photo are saved in “.txt” format files in the “calib” folder.

Get more details here.

Documentation

ML_Peach_RGB-D_Dataset_README.pdf

Demo Dataset

ML Peach RGB-D Demo Dataset.7z (226.7 MB)

Full Dataset

ML Peach RGB-D Dataset.7z (4.51 GB)

Decompression Key

The key will be provided soon. Please be patient. You can download the demo dataset in the meantime.