Spoiled and fresh fruit inspection dataset

The design of quality control systems in food has become an essential element in research to ensure a state suitable for consumption. It is necessary to develop automatic and efficient systems that can verify their condition before distribution. The proposed dataset can be used with a deep learning-based algorithm for the identification of the fruits and the state in which they are.

The data set has 8 different fruits:
-banana
-lemon
-lulo
-mango
-orange
-strawberry
-tamarillo
-tomato.

Two thousand images of each one of the types of fruits are acquired for a total of 16000 samples. Half of them correspond to fresh fruits and the other half to non-fresh or spoiled fruits.

For the acquisition of the dataset, changing of backgrounds, rotation of the fruits, distance of capture, and light variations were made, in order to make it robust.

食品质量控制系统的设计，已成为保障食品适宜食用状态的研究核心要素。开发可在食品流通前核验其状态的自动化高效系统，已成为必要之举。本数据集可配合基于深度学习的算法，用于识别水果种类及其所处状态。

本数据集涵盖8类不同水果：香蕉、柠檬、卢洛果（lulo）、芒果、橙子、草莓、树番茄（tamarillo）、番茄。每类水果均采集2000张图像，总计16000个样本。其中一半为新鲜水果样本，另一半为非新鲜或已变质的水果样本。

为提升数据集的鲁棒性，数据采集过程中对背景、水果摆放角度、拍摄距离及光照条件均进行了调整。