Mango and Banana Dataset (Ripe Unripe) : Indian RGB image datasets for YOLO object detection

'Mango and Banana Dataset (Ripe Unripe)' is the RGB image dataset. This dataset of 5000 photos of bananas and mangoes focuses on identifying ripe and unripe fruits. Each photograph has metadata that identifies whether or not the banana in the image is considered ripe. The data set was gathered in indoor as well as outdoor lighting conditions, to identify ripe and unripe Bananas and Mangoes. Each image in this dataset has a YOLO.txt label attached to it. This data can be used to train all YOLO Object Detection models. The dataset has been divided into two sections: Train and Test each of which contains 80% and 20% of the total data. Train folder contains 4000 images with labels and Test folder contains 1000 images with labels. The purpose of collecting this dataset was to create 'Ripe Unripe Fruit Detection System' using YOLOv8 Object detection model. Dimensions of image : 640 x 480

「芒果与香蕉数据集（成熟/未成熟）」为一套RGB图像数据集。该数据集包含5000张香蕉与芒果的图像，旨在实现成熟与未成熟水果的识别。每张图像均附带元数据，用以标注图像内的香蕉是否处于成熟状态。该数据集采集自室内与室外两种光照环境，以完成香蕉与芒果的成熟度识别任务。本数据集内的每张图像均附带对应的YOLO.txt格式标注文件，可用于训练各类YOLO目标检测模型。该数据集已划分为训练集与测试集两个子集，分别包含总数据量的80%与20%：训练集文件夹包含4000张带标注的图像，测试集文件夹则包含1000张带标注的图像。采集该数据集的初衷为基于YOLOv8目标检测模型搭建「成熟/未成熟水果检测系统」。图像分辨率为640×480。

搜集汇总

背景与挑战

背景概述

该数据集是一个包含5000张芒果和香蕉RGB图像的数据集，专门用于YOLO目标检测模型训练，重点识别水果的成熟与未成熟状态。数据集在室内外多种光照条件下采集，每张图像均附带YOLO格式标签，并已按80:20比例划分为训练集和测试集，图像尺寸统一为640x480，旨在支持成熟度检测系统的开发。