/Fruit-recognizer-application

Mobile application that can recognize a kind of fruit on a photo

Primary LanguageJupyter NotebookMIT LicenseMIT

Fruit_recognizer_application

version 1

Service for fruit recognition in the photo. Current version is able to classify 90 kinds of fruits. There is list of all fruits in the end of file.

Dataset - "Fruits 360 dataset" from kaggle: https://www.kaggle.com/moltean/fruits

There are 2 ways to run server and classificator:

  1. Without using Docker:

    1. Install python 3.6 interpreter (64-bit)

    2. Install all packages from requirements.txt To do this use command in command line: pip install -r requirements.txt

    3. Run server.py file to start the server

    4. Run send_picture_for_checking.py script to check the work of classificator. It send picture to server and print classification result.

  2. Using Docker:

    1. From /root directory run command to build Docker image: sudo docker build -t recognitron
    2. From /root directory run command to run Docker image in a container: sudo docker run -d -p 4000:80 recognitron

There are 2 ways to create Android Debug APK:

  1. Using Docker (NOTE: needs quite a lot of time and stable internet connection to build APK, although nothing beside the Docker is needed to build):

    1. Set permition to execute for client/gradlew script (sudo chmod +x gradlew).
    2. From /client directory run command to build Docker image: sudo docker build . -t fruit-recognizer
    3. From /client directory run command to run Docker image in a container: sudo docker run --mount type=bind,source="$(pwd)/..",target=/app fruit-recognizer NOTE: Yes, you need to bind mount the whole project, with server, because code in client depends on .git directory - this will be fixed as soon as client will reside in its own repo.
    4. Generated APK is located under client/app/build/outputs/apk/debug

  2. Using Gradle (NOTE: JDK8 and Android SDK 28 with Android Build Tools 28.0.3 are needed in order to build APK):

    1. Set permition to execute for client/gradlew script (sudo chmod +x .gradlew).
    2. From /client directory run command to build APK: sudo ./gradlew clean build on Linux, or gradlew clean build on Windows
    3. Generated APK is located under client/app/build/outputs/apk/debug

Classifier model

A convolutional neural network is used to classify. All CNN architectures are in models_development folder. Server use models package, which downloads model architecture (as .json) and weights(as .hdf5) from Google Drive (during the first run).

All kind of fruits:

'Apple Braeburn', 'Apple Golden 1', 'Apple Golden 2', 'Apple Golden 3', 'Apple Granny Smith', 'Apple Red 1', 'Apple Red 2', 'Apple Red 3', 'Apple Red Delicious', 'Apple Red Yellow 1', 'Apple Red Yellow 2', 'Apricot', 'Avocado', 'Avocado ripe', 'Banana', 'Banana Lady Finger', 'Banana Red', 'Cactus fruit', 'Cantaloupe 1', 'Cantaloupe 2', 'Carambula', 'Cherry 1', 'Cherry 2', 'Cherry Rainier', 'Cherry Wax Black', 'Cherry Wax Red', 'Cherry Wax Yellow', 'Chestnut', 'Clementine', 'Cocos', 'Dates', 'Granadilla', 'Grape Blue', 'Grape Pink', 'Grape White', 'Grape White 2', 'Grape White 3', 'Grape White 4', 'Grapefruit Pink', 'Grapefruit White', 'Guava', 'Huckleberry', 'Kaki', 'Kiwi', 'Kumquats', 'Lemon', 'Lemon Meyer', 'Limes', 'Lychee', 'Mandarine', 'Mango', 'Mangostan', 'Maracuja', 'Melon Piel de Sapo', 'Mulberry', 'Nectarine', 'Orange', 'Papaya', 'Passion Fruit', 'Peach', 'Peach 2', 'Peach Flat', 'Pear', 'Pear Abate', 'Pear Monster', 'Pear Williams', 'Pepino', 'Physalis', 'Physalis with Husk', 'Pineapple', 'Pineapple Mini', 'Pitahaya Red', 'Plum', 'Pomegranate', 'Quince', 'Rambutan', 'Raspberry', 'Redcurrant', 'Salak', 'Strawberry', 'Strawberry Wedge', 'Tamarillo', 'Tangelo', 'Tomato 1', 'Tomato 2', 'Tomato 3', 'Tomato 4', 'Tomato Cherry Red', 'Tomato Maroon', 'Walnut'