蔬菜图像集

从蔬菜生产到交付，有几个常见步骤是手动操作的。比如采摘和分类蔬菜。因此，我们决定通过开发一个可以检测和分类蔬菜的模型，使用深度神经架构来解决这个问题。该模型可以在不同类型的设备中实现，也可以解决与蔬菜识别相关的其他问题，例如自动标记蔬菜，而无需任何人工操作。

Several common steps from vegetable production to delivery are performed manually. For example, vegetable harvesting and sorting are typical manual operations. To address this problem, we developed a model for vegetable detection and classification using deep neural network architectures. This model can be deployed across various types of devices, and it can also tackle other issues related to vegetable recognition, such as automatically labeling vegetables without any manual intervention.

搜集汇总

背景与挑战

背景概述

该数据集包含15类常见蔬菜的21000张图像，每类1400张，图像尺寸为224×224，格式为.jpg。数据集已划分为训练集（70%）、验证集（15%）和测试集（15%），并提供了生成元数据文件的Python脚本。