FOOD PRODUCTS INGREDIENT DATASET

Ingredients are the building blocks of packaged food product that enables a product to have texture, flavor, and nutritional information as well. The companies dealing with these products add some ingredients in the form of artificial flavor enhancers, color, and sweeteners to enhance the flavor and appearance of a product. Artificial additives may have health implications as the consumers are not very much aware of the products that are being used therefore it is a complex task for a common person to figure out manually that the product based on its ingredients is good for health. Complex tasks are very efficiently performed in the field of artificial intelligence and machine learning as they are trained for solving complex mathematical problems. With the availability of an appropriate dataset, machine learning can easily solve complex tasks using statistical techniques by model built on available dataset. To achieve the objective, the ingredients dataset has generated by extracting text from images containing ingredient lists of different ready-to-eat food products. In the dataset, several images of different lines of food products have been acquired to create a harmonized dataset. The dataset contains the list of ingredients named with 7 categories and corresponding attributes with appropriate labeling. The dataset will be very useful for training and testing machine learning models for food product classification through their ingredients.

配料是预包装食品的核心构成要素，同时赋予产品特定的质地、风味与营养属性。相关食品生产企业常会添加人工风味增强剂、着色剂与甜味剂等配料，以优化产品的风味与外观表现。由于普通消费者对所使用的添加剂成分普遍缺乏认知，人工添加剂可能存在健康隐患，因此普通人仅凭手动核查配料列表来判断食品是否健康，是一项极具挑战性的复杂任务。人工智能与机器学习领域经过复杂数学问题求解训练，可高效处理各类复杂任务。借助适配的数据集，机器学习模型可通过统计技术基于数据集构建，从而轻松解决这类复杂问题。为实现上述目标，本研究通过提取各类即食食品配料列表图像中的文本信息，构建了该配料数据集。数据集采集了不同产品线食品的多幅图像，以构建标准化统一的数据集。该数据集涵盖7大类命名的配料清单，以及带有规范标注的对应属性信息。本数据集可有效用于基于配料信息开展食品分类任务的机器学习模型训练与测试工作。

搜集汇总

背景与挑战

背景概述

该数据集是一个包装食品成分数据集，包含从食品成分图片中提取的文本信息，涵盖7个类别的成分名称和属性，适用于机器学习模型训练和食品分类任务。