Forager App

Forager is an iOS application designed to identify plants using photos. It leverages CoreML for machine learning-based classification and provides users with plant names and descriptions from a predefined dictionary.

---

Prerequisites

Hardware

• An iPhone to run the app.

Software

• Xcode installed on your macOS system.
• Access to the Forager project code repository.

---

Setup Instructions

1. Pull the Code

   • Clone the Forager app repository from GitHub to your local machine.

2. Open the Project

   • Launch Xcode and open the Forager project.

3. Connect an iPhone

   • Plug in your iPhone to the computer via USB.
   • Follow Apple's instructions to enable your iPhone as the run destination in Xcode.

4. Select Destination

   • In Xcode, select your iPhone as the destination device by clicking the device selection dropdown.

5. Run the App

   • Click the "Run" button (triangle icon in the corner of Xcode).
   • The app will be installed and launched on your iPhone.

---

Using the Forager App

1. Launch the App

   • Open the app on your iPhone.
   • During the loading screen, the app will load plant data from a CSV file to create a plant dictionary for identification.

2. Take or Select a Photo

   • Press the "Take a Photo" button on the app.
   • Either take a photo of a plant using the camera or select an existing photo from your gallery. The app uses UIImagePickerController to handle this process.

3. Plant Identification

   • The app processes the image using a trained CoreML model.
   • The image is converted into a CVPixelBuffer, which is passed to the model for classification.
   • The model compares the prediction against the plant dictionary:
     • If the plant is identified, the app displays its name and description.
     • If the plant is not in the dictionary, the app displays a message indicating no information is available.

---

Known Bugs and Limitations

1. Confidence Score

   • The app always returns a prediction, even if the confidence score is low.
   • Future Improvement: Set a minimum confidence threshold to identify plants accurately.

2. Model Errors

   • The machine learning model may misidentify plants due to limitations in training or data quality.

---

Technical Notes

1. Core Features

   • The app uses SwiftUI for UI design and CoreML for machine learning integration.
   • Key imports include:
     • UIKit
     • Foundation
     • Vision
     • XCTest
     • CoreML

2. Plant Data

   • Plant data is stored in a CSV file, which is loaded during app initialization to build a plant dictionary.

3. Predictions

   • The app processes images using pixel data and matches predictions to its dictionary for results.

4. Testing

   • Test the app by taking or selecting photos of plants.
   • Validate the app's ability to identify plants using its dictionary and note any misclassifications or issues.

---

Future Enhancements

• Implement a confidence threshold for predictions to improve identification accuracy.
• Enhance the CoreML model with additional training data for better predictions.

Time Log

Date	Activity	Members Present	Total Hours
September 27-29	Met multiple times to discuss project ideas.	Anthony, Wyatt, Dylan, Mia, Jake	2 hours
October 28	Met after class to break up tasks.	Anthony, Wyatt, Dylan, Mia, Jake	1 hour
November 13	Made a group presentation and discussed final tasks.	Anthony, Wyatt, Dylan, Mia, Jake	2 hours
Total Hours			5 hours

---

Disclaimer

The total hours listed reflect the group meeting times and do not represent the individual time commitment or effort of each group member.