An android application, written in Kotlin, that uses the Clarifai API to identify images of fruits and hosts educational pages geared towards children.
Report Bug
·
Request Feature
This project was developed as a part of the WiCSE @ UF Shadowing Program during Fall 2020. Under the guidance of a corporate mentor from Traject, I had to develop an application over the course of the semester.
Tasks:
- Sketch wireframe to plan app structure
- Collaborate with student artist to develop the UI
- Create a timeline for app milestones
- Write reports outlining project progress
- Present project results to mentors and peers
- Research and learn about Kotlin and Android development
- Develop and test my android application
- Maintain time logs of my activities
- Host biweekly meetings with the mentor
The goal of the project was to develop a mobile application that is designed to teach children about fruits. The app takes a photo of a fruit, and then sends the image to the Clarifai API. There, an image recognition model will identify the fruit, and then return a label. The app then uses the label to redirect the user to the fruit's respective page, with educational information about that fruit. Educational information will include color, pronunciation, and origin. UI/UX elements, such as vivid palettes and friendly interface, will be emphasized in order to accommodate the attention of young children.
- Apple
- Avocado
- Banana
- Blueberry
- Cherry
- Coconut
- Dragonfruit
- Durian
- Grape
- Grapefruit
- Guava
- Kiwi
- Lemon
- Lime
- Lychee
- Mango
- Orange
- Papaya
- Passionfruit
- Peach
- Pear
- Persimmon
- Pineapple
- Pomegranate
- Raspberry
- Strawberry
- Watermelon
- Install Android Studio.
- Make Clarifai account, create a project, and make an API Key.
- Install and Configure NDK and CMake in Android Studio. Tutorial
- Create NDK file in order to securely store API keys. Tutorial
- Android Studio
- Clarifai Account and API Secret + ID
- Android Device is required, as the app makes camera intents.
- Take Picture of Fruit:
- Educational pages about respective fruit:
- App layout:
- Child-Friendly UI with hand-drawn images
- Image Recognition and Identification of 25+ Fruits
- Fruit origin images are publicly hosted on AWS S3 bucket
- Sends camera intent to take images
- Alert Dialog and halt app if no internet access
- Asynchronous code to send gRPC calls and load images from URLs
- App saves state if user leaves app
- Educational pages that teaches color, name, and origin for each fruit
- App has option to save image to phone gallery
- Loading photos in app is optimized by scaling the bitmap
- Cache is cleaned upon app reinitialization
- Tests to verify app has proper navigation
Raymond Yu (Developer) - wrg7yu@gmail.com
Alana Jones (Artist) - alanaj829@gmail.com