This repository contains an Android app built with Unity 6, using Barracuda inference and OpenCV for detecting and classifying trash items, particularly optimized for identifying plastics. The app leverages Unity 6 with Barracuda inference and OpenCV for image processing.

• Unity 6
• Barracuda Inference (for Unity)
• OpenCV

1. Clone the Repository

   git clone https://github.com/Kilo-zolo/gotcha_trash.git

2. Install Unity Hub and Unity 6

   • Download and install Unity Hub.
   • Ensure Unity 6 is installed via Unity Hub.

3. Open Project in Unity

   • Open Unity Hub.
   • Click on Open Project and select Add project from disk.
   • Choose the cloned repository directory.

• Simply click the Play button at the top-center of Unity to run within the editor.

Step 1: Android Device Setup

• Enable developer mode on your Android device:
  • Enable Developer Mode
• Turn on USB Debugging (see above video for step-by-step).

• Go to File > Build Settings > Player Settings.
• Under the Android tab:
  • Uncheck Auto Graphics API.
  • Ensure API level 29 (Android 10) or higher is selected for minimum API level.
• Connect your Android device via USB.
• Click on Build and Run.

Compiling might take a minute, so please be patient.

• Watch this video for additional guidance.

• Point your camera at objects.
• The app will detect trash items (currently optimized for densely packed plastic items).
• Enjoy exploring the object detection capabilities!

• Ensure Unity and Android SDKs are properly set up before building.
• Compilation may take a few moments.