STEM 連繫 (S1–S3)STEM connection (S1–S3)實驗儀器的影像識別程式Image recognition programme for laboratory apparatus
學習技能：Learning skill:
編程訓練Programming
前言Introduction
在課本的1.4A及1.4B章節，學生已認識實驗室中一些常見的儀器，並學懂繪畫儀器的縱切面圖。在這個編程活動中，學生將學習如何製作和「訓練」一個影像識別程式，以識別用戶所繪畫的縱切面圖。
In Section 1.4A and 1.4B, students have studied some common apparatus in the laboratory and have also learned how to draw the vertical section diagrams of the apparatus. In this coding activity, students will learn how to create and ‘train’ an image recognition programme to recognize a vertical section diagram drawn by a user.
儀器的縱切面圖Vertical section diagrams
儀器的縱切面圖可幫助我們清楚顯示實驗裝置。縱切面圖顯示儀器從側面觀察的樣子，以下是一些例子。
Vertical section diagrams help us show experimental set-up clearly. These diagrams show how apparatus looks from the side. Here are some examples.
試管
test tube
大試管連膠塞
boiling tube with rubber stopper
燒杯
beaker
錐形燒瓶
conical flask
量筒
measuring cylinder
錶面玻璃
watch glass
蒸發皿
evaporating dish
支架連夾
stand and clamp
集氣瓶
gas jar
玻璃棒
glass rod
漏斗連濾紙
filter funnel with filter paper
滴管
dropper
影像識別技術Image recognition technology
甚麼是影像識別程式？What is an image recognition programme?
影像識別程式是採用人工智能（AI）的一種軟件，可用來分辨數碼圖像中的物體或特徵。程式會使用機器學習的演算法來分析圖像，並與存於資料庫的已知圖像進行比較，以分辨圖像中的內容。透過提供更大的圖像資料庫，程式可識別大量圖像，並經過「訓練」來提高識別物體和特徵的能力和準確度。在識別出物體或特徵後，程式可向使用者或其他連接的程式或設備提供適當的反應。
影像識別技術已日漸廣泛應用於日常生活，例如自動駕駛、人臉識別和保安系統。
An image recognition programme is a type of artificial intelligence (AI) software that can identify objects or features in digital images. The programme uses machine learning algorithms to analyze an image and compare it to a database of known images in order to identify what is in the image. By analysing a massive quantity of images provided in larger datasets, the programme can be ‘trained’ to improve its accuracy and ability in recognizing objects and features. After recognizing an object or feature, the programme can give the corresponding responses to the users or other linked programmes or equipment.
The use of image recognition technology is now widespread in various applications such as autonomous vehicles, facial recognition, and security systems.
影像識別技術可用於自動駕駛的車輛
Image recognition technology can be used on autonomous vehicles.
Image by macrovector on Freepik
編程活動Coding activity
使用Scratch來製作簡單的影像識別程式Using Scratch to build a simple image recognition programme
Scratch 是一個網上電腦程式開發平台，讓學生學習編寫電腦程式、遊戲、動畫或互動程式等。請參考以下步驟，學習如何使用Scratch製作一個簡單的影像識別程式。
Scratch is an online computer programming platform that enables students to create interactive stories, games, and animations. Follow the steps below to learn how to use Scratch to create a simple image recognition programme.
第一部分—開啟程式Part I - Opening the programme
-
點擊以下按鈕下載或檢視示範專案。
示範專案：
下載
檢視
Click on the following buttons to download or view the demonstration project.
Demonstration project:
Download
View
-
如在網上檢視專案，你可以點擊右上方的「切換到程式頁面 (See inside)」按鈕，查看這程式是如何製作的。
If you are viewing the project online, you can click on the ‘See inside （切換到程式頁面）’ button in the top right corner to see how it is built.
第二部分—訓練程式來識別圖像Part II - Training the programme to recognize a drawing
-
點擊「綠旗」按鈕來啟動程式。
Click on the ‘green flag’ button to start the programme.
-
依照畫面上的指示分別繪畫試管和燒杯的縱切面圖。
Follow the on-screen instructions to draw vertical section diagrams of a test tube and a beaker, respectively.
-
程式將「學習」你繪畫的圖案，用於以後分辨新的繪圖。
The programme will then ‘learn’ the pattern of your drawings and use it to identify new drawings later.
第三部分—測試程式並給予回應Part III - Testing the programme and giving feedback
-
依照畫面上的指示，繪畫一個試管或燒杯的縱切面圖。
Follow the on-screen instructions to test the programme by drawing the vertical section diagram of a test tube or a beaker.
-
程式將會分辨你所繪畫的實驗儀器。你可提供回饋給程式，讓它知道分辨是否正確。回饋可幫助程式提高識別圖像的能力和準確度。
The programme will identify the apparatus that you have drawn. Provide feedback to the programme to let it know whether its identification result is correct or not. Your feedback will help the programme improve its accuracy and ability to recognize drawings.
第四部分—訓練和修改你自己的影像識別程式Part IV - Training and modifying your own image recognition programme
-
點擊以下按鈕下載或檢視延伸專案。
延伸專案：
下載
檢視
Click on the following buttons to download or view the extended sample project.
Extended project:
Download
View
-
在這個專案中，你可以繪畫任何圖案，並告訴程式你所繪畫的是甚麼。再依照畫面上的指示，測試程式是否能夠成功識別你的繪圖。
In this project, you can create any drawings and tell the programme what you have drawn. Follow the instructions again to test whether the programme can successfully identify your drawings.
-
你還可以自行修改專案，製作一個獨一無二的影像識別程式，例如更改程式中角色的設計和反應。
You can also modify the project with your own ideas to create a unique image recognition machine, such as changing the design and response of the character in the programme.
討論Discussion
影像識別技術在科學上的應用The use of image recognition technology in science
影像識別技術可以應用於科學上不同的範疇。科學家借助這項技術，能快速和準確地分析大量數據，從而推動對人體、自然界和宇宙的發現和理解。
例如，在醫學上，影像識別技術可以用於分析細胞和器官的影像，以分辨其結構或組成部分有沒有異常或變化。這有助科學家更好地理解疾病並開發新的治療方法。
試從互聯網上搜尋資料，並與同學討論影像識別技術在不同科學範疇上的其他應用。
Image recognition technology can be used in various branches of science. By using this technology, scientists can analyze large amounts of data quickly and accurately for making new discoveries that can advance their understanding of our body, the natural world and the universe.
For example, in medicine, image recognition technology can be used to analyze images of cells and organs to identify abnormalities or changes in their structure or composition. This helps scientists better understand diseases and develop new treatments.
Try to search from the Internet and discuss with your classmates the other uses of image recognition technology in different branches of science.
選擇一項科學範疇：
物理學／化學／生物學／天文學／地質學／醫學／其他：__________
影像識別技術在這科學範疇上可以如何應用？
Select a branch of science:
Physics / Chemistry / Biology / Astronomy / Geology / Medicine / Others:__________
How image recognition technology can be used in this branch of science?