Object Recognition App Using Look Extension in MIT App Inventor

on

 



IMAGE CLASSIFICATION

Want to build an AI-powered image recognition app without training your own model? With the Look Extension in MIT App Inventor, you can classify objects in images using a pre-trained Google image classifier — no machine learning training required!

 1️⃣ What the Look Extension Does

The Look Extension can:

  • Identify objects and scenes in pictures

  • Use a pre-trained Google image classification model

  • Return labels with confidence scores

Example:

 You take a picture →
The extension returns results like:

  • Dog (92% confidence)

  • Tree (78% confidence)

  • Car (65% confidence)

This makes it perfect for beginners exploring Artificial Intelligence in mobile apps.

2️⃣ What You’ll Need

Before starting, make sure you have:

  • A free account on MIT App Inventor

  • The Look Extension (.aix file) (Available in the AI & ML section of MIT App Inventor Extensions)

  • Internet connection (required for classification)

  • Basic knowledge of buttons and labels

 3️⃣ Add Components (Designer Tab Setup)

 Step 1: Import the Look Extension

  1. Go to Extensions

  2. Click Import Extension

  3. Upload the downloaded .aix file

  4. The Look extension will appear in the Extensions drawer

 Step 2: Add Required Components

Add the following components:

ComponentPurpose
Button1“Take Picture”
Camera1Capture image
Image1Display captured image
Label1Show classification results
WebViewer1Optional – View image or result page
Classify ButtonTrigger image recognition
Toggle ButtonSwitch between images
Look1AI classification extension

How the App Works (Logic Flow)

Step 1 – Capture Image

When Take Picture Button is clicked:

  • Call Camera1.TakePicture

When Camera1.AfterPicture:

  • Set Image1.Picture to the captured image

Step 2 – Classify Image

When Classify Button Clicked:

  • Call Look1.ClassifyImage

  • Pass the image path

Step 3 – Display Results

When Look1.GotClassification:

  • Get label list

  • Get confidence scores

  • Set Label1.Text to show results

Example output:

Dog – 92%
Animal – 88%
Pet – 80%

Step 4 – Toggle Button (Optional Feature)

The Toggle button can:

  • Switch between gallery image and captured image

  • Enable/Disable classification mode

This adds interactive control to your AI app.

 Key Concepts Students Learn

✅ Artificial Intelligence basics
✅ Image classification
✅ Using Extensions in MIT App Inventor
✅ Event-driven programming
✅ Working with lists and data results
✅ Building real-world AI apps

Why This Project Is Great for Students

  • No machine learning training required

  • Beginner-friendly

  • Real AI integration

  • Practical application of computer vision

  • Ideal for Grades 6–10

Possible Enhancements

You can extend this app by adding:

  • Voice output (Text-to-Speech)

  • Save classification history

  • Confidence percentage formatting

  • Display top 3 results only

  • Upload image from gallery

  • Create a plant/animal recognition app

 Applications

  • Smart object detection app

  • Educational AI demo

  • Science fair project

  • Beginner AI portfolio project

  • Image recognition experiment


CLICK BELOW LINK TO DOWNLOAD CODE:


DOWNLOAD EXTENSION:

DOWNLOAD EXTENSION


Comments

Post a Comment