基于ESP32与CNN的卡路里识别相机设计与应用

Design and Application of a Calorie Recognition Camera Based on ESP32 and CNN

摘要: 基于ESP32微控制器与卷积神经网络(CNN)，设计并实现了一种轻量化的卡路里识别相机。系统通过低功耗摄像头实时采集餐食图像，利用ESP32进行初步预处理后传输至云端，由优化后的CNN模型完成食物识别与卡路里计算。本研究不仅实现了饮食管理的便携化与高效化，还具备教育推广价值，可作为人工智能与嵌入式教学的教具载体，为学生提供实践平台，促进AI知识在健康管理与教育中的融合应用。

Abstract: A lightweight calorie recognition camera is designed and implemented based on the ESP32 microcontroller and convolutional neural networks (CNN). The system captures food images using a low-power camera, preprocesses them on ESP32, and sends them to the cloud, where an optimized CNN model performs food recognition and calorie estimation. This study not only enables portable and efficient dietary management but also provides educational value as a teaching tool for AI and embedded systems, offering students a practical platform and promoting the integration of AI knowledge into health management and education.