Oranges Classification

This dataset containg 1900+ images divided into fresh oranges and rotten oranges. In an orange packing factory, a video was recorded, by placing the camera parallel and above the oranges conveyor. The video was captured for 10 minutes with a quality of Ultra High Definition (4K) with 60 frames per second and a High Dynamic Range feature. The video was changed from High Dynamic Range to Standard Dynamic Range by the use of Splice - Video Editor & Maker software. The video is inserted to developed algorithm operating video processing on it and creating the frames. Resulting in 1,743 fresh oranges and 245 rotten oranges.本数据集包含1900余张图像，分为新鲜橙子与腐烂橙子两类。数据采集自橙子包装工厂：将摄像头平行放置于橙子传送带的上方，录制了一段时长10分钟的视频。该视频采用超高清（Ultra High Definition, 4K）规格录制，帧率为60帧每秒，并搭载高动态范围（High Dynamic Range, HDR）功能。随后使用Splice - Video Editor & Maker软件，将该视频从高动态范围格式转换为标准动态范围（Standard Dynamic Range）格式。将转换后的视频输入至自研的视频处理算法中，经算法处理并提取帧图像后，最终得到1743张新鲜橙子图像与245张腐烂橙子图像。

搜集汇总

背景与挑战

背景概述

该数据集是一个用于橙子分类的图像数据集，包含超过1900张图像，分为新鲜橙子（1,743张）和腐烂橙子（245张）两类，适用于计算机视觉和机器学习任务。数据来源于橙子包装工厂的传送带视频，以4K超高清拍摄，并经过处理生成图像，旨在支持水果质量控制和工业4.0应用。