尿液細菌與微生物 AI 影像辨識學習教材

產品說明

Product Description

本教學套組以尿液微生物與細菌 52 張 2D 學習卡牌與 8 種 3D 立體模型為核心教具，搭配羅技高解析鏡頭模組，分別以 Teachable Machine 進行 2D 卡牌影像分類或以 Roboflow AI 平台進行 3D 模型物件偵測學習，引導學習者完成鏡頭連接、影像擷取、資料標註、模型訓練與驗證測試之完整 AI 影像辨識學習流程。

學習者更可透過 Vibe Coding 方式，以自然語言提示驅動 AI 輔助程式生成，串接 Roboflow 前後台 API，自行開發個人化操作介面，將抽象的顯微世界轉化為可動手操作的 AI 應用實作體驗。

This educational kit centers on 52 urinary microorganism and bacteria 2D learning cards and 8 types of 3D physical models, paired with a Logitech high-resolution camera module. Learners use Teachable Machine for 2D card image classification or the Roboflow AI platform for 3D model object detection, completing a full AI image recognition workflow that includes camera setup, image capture, data annotation, model training, and validation testing.

Through Vibe Coding, learners use natural-language prompts to drive AI-assisted code generation, connecting to Roboflow's front-end and back-end APIs to independently build personalized interfaces — transforming the abstract microscopic world into a hands-on AI application experience.

核心功能

Core Capabilities

互動式學習Interactive Learning

影像資料集建立Image Dataset Building

AI 即時辨識AI Real-Time Recognition

尿液微生物與細菌模型Urinary Microorganism & Bacteria Models

本套組整合具象化教學設計與 AI 工具，以 2D 卡牌與 3D 立體模型呈現肉眼難見的微觀物件，帶領學習者從建立影像資料集、訓練辨識模型，到串接 API、開發互動介面，完整 AI 影像辨識的開發流程。

結合自然語言生成的輔助程式設計，學習者無需深厚程式背景也能自主學習，真正實現跨領域 AI 學習體驗。

This kit combines concretization-based teaching with AI tools, using 2D cards and 3D models to reveal microscopic objects invisible to the naked eye. Learners follow the complete AI image-recognition workflow — from building datasets and training models to connecting APIs and developing interactive interfaces.

With natural-language-assisted coding, learners can study independently even without a deep programming background, achieving a true cross-disciplinary AI learning experience.

細菌 2D 學習卡牌

Bacteria 2D Cards [ Compatible with Teachable Machine ]

以實體卡牌於 Teachable Machine 建立影像分類模型，快速完成訓練與即時辨識。介面簡單直覺、無需寫程式，學習者即可快速掌握影像分類的基礎邏輯。

Using physical cards, learners build an image classification model in Teachable Machine for fast training and real-time recognition. The interface is simple and code-free, helping learners quickly grasp the basics of image classification.

細菌 3D 立體模型

Bacteria 3D Models [ Compatible with Roboflow ]

以鏡頭拍攝立體模型，於 Roboflow 標註訓練物件偵測，進行即時辨識。相較影像分類，物件偵測需同時判斷類別與位置，體驗更貼近實務的 AI 視覺辨識流程。

Using the camera, learners photograph 3D models and annotate them in Roboflow to train a detection model for real-time recognition — identifying both category and location for AI vision.

學習流程

AI Learning Workflow

1

將鏡頭連接電腦Connect Camera to Computer

以 USB 將辨識鏡頭模組接上電腦，確認瀏覽器已取得攝影機權限。

Connect the recognition camera module to your computer via USB, and confirm the browser has camera access.
2

影像擷取Image Capture

拍攝 2D 學習卡牌與 3D 立體模型，建立測試影像資料集。

Photograph the 2D learning cards and 3D models to build a test image dataset.
3

分類與標註Classification & Annotation

2D 卡牌於 Teachable Machine 建置類別；3D 模型於 Roboflow 框選標註。

Set up categories for 2D cards in Teachable Machine; annotate 3D models with bounding boxes in Roboflow.
4

模型訓練Model Training

訓練影像辨識模型，觀察準確率與訓練成效。

Train the image recognition model and monitor accuracy and training performance.
5

辨識測試Recognition Testing

即時辨識卡牌與實體模型，觀察辨識結果與準確率。

Test real-time recognition on cards and physical models, and review the results and accuracy.
6

網站開發Website Development

於 Visual Studio Code 中以 Vibe Coding 方式串接辨識模型，開發即時辨識網站。

Using Vibe Coding in Visual Studio Code, connect the recognition model and build a real-time recognition website.

學習特色

Key Features

52 種尿液微生物與細菌 2D 卡牌

52 Types of Urinary Microorganism & Bacteria 2D Cards
8 種尿液微生物與細菌 3D 模型

8 Types of Urinary Microorganism & Bacteria 3D Models
搭配基礎 Teachable Machine AI 影像辨識教學

Basic AI Image Recognition Training with Teachable Machine
搭配高階 Roboflow AI 影像辨識教學

Advanced AI Image Recognition Training with Roboflow
結合 Vibe Coding 自然語言生成

Natural-Language Code Generation with Vibe Coding
附羅技高解析鏡頭模組

Includes Logitech High-Resolution Camera Module

本套組專為教育訓練目的設計，所有模型與辨識結果不具臨床診斷效力，不得作為醫療依據。

For educational and training purposes only. All models and recognition outputs do not constitute clinical diagnostic results and must not be used as medical references.