Ice crystal image dataset from CPI, 2DS and HVPS probes

For conducting a study on cloud particle shape recognition, a dataset comprising a sufficient number of accurately labeled particles is essential. Our dataset of particle images is derived from several field observation projects that run in China and USA. These cloud particle images were recorded by the airborne probes - cloud particle imager (CPI), two-dimensional Stereo (2D-S) Probe and high volume particle spectrometer (HVPS). The particle images were classified into nine categories: plate (Pla), rosette (Ros), sphere (Sph), column (Col), column aggregate (Col. Agg.), dendrite (Den), graupel (Gra), aggregate (Agg), and irregular (Irr).The dataset contains 8,100 representative images of particles, with 300 images for each of the different habit classes for each probe. This dataset were randomly divided into a training set and a test set with a ratio of 4:!.