# DeepPod: A convolutional Neural Network Based Quantification of Fruit Number in Arabidopsis

# Azam Hamidinekoo, Gina A. Garzon-Martinez, Morteza Ghahremani, Fiona M. K. Corke, Reyer Zwiggelaar, John H Doonan, and Chuan Lu

## Plant Specie used in this study

Arabidopsis thaliana inflorescences

## Types of Data

There are two folders with the datasets used to test and establish the CNN pipeline.\

### Directories:

#### Set_1

Images used for manual annotation and training the convolutional neural network. Each folder corresponds to a different individual of Arabidopsis thaliana. They 
contain the image in *.png format and four *.txt files with the manual annotations of the image. In each image file, the first three numbers after the name of 
the experiment (eg. AT0XX_152XXX) corresponds to the RIL number as described by Kover et al. 2009. 

#### Set_2

Images used for testing the performance of the model. Each folder corresponds to a different individual of Arabidopsis thaliana. They contain images in *.png format.

### Files:

#### Set_1

A *.csv file contains the manual counting data of Set_1.

#### Set_2

A *.csv file contains the manual counting data of Set_2.