Manuscript Title
Identification of grass pollen through the quantitative analysis of surface ornamentation and texture

Author Names
Luke Mander, Mao Li, Washington Mio, Charless C. Fowlkes, and Surangi W. Punyasena  

Readme
There are seven items in this repository that correspond to Datasets S1, S2 and S3 in the manuscript entitled “Identification of grass pollen through the quantitative analysis of surface ornamentation and texture” by Luke Mander, Mao Li, Washington Mio, Charless C. Fowlkes, and Surangi W. Punyasena. These items are images of the pollen of 12 species of grasses. These species are from three subfamilies within Poaceae, the Pooideae, the Chloridoideae and the Panicoideae. The images were acquired using scanning electron microscopy (SEM). Twenty pollen grains of each species were imaged. Each specimen was imaged once at x2,000 magnification, once at x6,000 magnification, and once at x12,000 magnification. Dataset S1 consists of raw SEM images and metadata from every specimen at x2,000, x6,000 and x12,000 magnification as provided by the instrument (a JEOL JSM-6060LV SEM running at 15kV). Dataset S1 has been split into three parts corresponding to: the Pooideae (Part 1), the Chloridoideae (Part 2), and the Panicoideae (Part 3). Dataset S2 consists of x6,000 x12,000 SEM images of each specimen of grass pollen cropped to 400x400 pixels. Dataset S2 has also been split into three parts: the Pooideae (Part 1), the Chloridoideae (Part 2), and the Panicoideae (Part 3). Dataset S3 consists of files used to test human ability to classify SEM images of grass pollen. These include a reference library, unclassified engraved images and master list of image classification.