tes

Dockerfile

@@ -1,17 +1,17 @@
-FROM python:3.8-slim-buster
-
-WORKDIR /app
-
-COPY requirements.txt requirements.txt
-
-RUN pip install --upgrade pip
-
-RUN pip install -r requirements.txt
-
-COPY . .
-
-ENV PYTHONUNBUFFERED=1
-
-EXPOSE 8080
-
-CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8080"]
+FROM python:3.8-slim-buster
+
+WORKDIR /app
+
+COPY requirements.txt requirements.txt
+
+RUN pip install --upgrade pip
+
+RUN pip install -r requirements.txt
+
+COPY . .
+
+ENV PYTHONUNBUFFERED=1
+
+EXPOSE 8080
+
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8080"]

README.md

@@ -0,0 +1,28 @@
+![TF](https://github.com/Mubazir-Bangkit-2023/mubazir-machine-learning/assets/95016158/cf4884a9-2a4d-4148-a5a7-24a57c009da0)
+
+# Food Image Classification
+Developing a deep learning model for food image classification using Tensorflow: Creating a Convolutional Neural Network (CNN) model to categorize fruits and vegetables with Tensorflow.
+
+## Project Overview
+The objective of this project is to construct a model capable of precisely categorizing images of fruits and vegetables into predetermined classes. This model has the potential to be incorporated into applications, enabling the automatic identification and classification of the fruit or vegetable type and determining its freshness based on images uploaded by users.
+
+## Dataset
+The dataset employed in this project comprises images showcasing a variety of fruits and vegetables systematically arranged into distinct categories. Every image is annotated with the specific category of the fruit or vegetable and its corresponding freshness classification.
+
+## Feature
+- Data Augmentation 
+- CNN (Convolutional Neural Networks)
+
+## Requirements 
+- Tensorflow 
+- Matplothlib
+- Numpy
+- Pillow
+- Scikit-learn
+
+## Results
+The model achieved a test accuracy of XX% on the test dataset. The training and validation loss/accuracy plots can be found here.
+
+## Future Work
+- Expand the data set to include a wider variety of fruits and vegetables
+- Explore advanced architectures and methodologies in experiments to enhance accuracy further.

app.py

@@ -1,61 +1,77 @@
-import os
-import uvicorn
-import numpy as np
-import traceback
-from io import BytesIO
-
-from pydantic import BaseModel
-from urllib.request import Request
-from fastapi import FastAPI, Response, UploadFile
-from utils import load_image_into_numpy_array
-from tensorflow.keras.models import load_model
-from tensorflow.keras.preprocessing.image import img_to_array
-from tensorflow.keras.preprocessing.image import load_img, img_to_array
-
-app = FastAPI()
-
-@app.post("/predict_image")
-async def predict_image(img: UploadFile, response: Response):
-    try:
-        # Checking if it's an image
-        if img.content_type not in ["image/jpeg", "image/png"]:
-            response.status_code = 400
-            return "File is Not an Image"
-
-        # Read file content
-        file_content = await img.read()
-
-        # Convert bytes to a file-like object
-        file_like_object = BytesIO(file_content)
-
-        # Load the image from the file-like object
-        img = load_img(file_like_object, target_size=(150, 150))
-
-        # Load the model
-        model = load_model('model.h5')
-
-        # Prepare the image for prediction
-        image = img_to_array(img)
-        image = np.expand_dims(image, axis=0)
-
-        # Predict the class of the image
-        arr = model.predict(image)
-
-        if arr[0][0]==1:
-            labels='Fresh Apples'
-        elif arr[0][1]==1:
-            labels='Fresh Banana'
-        elif arr[0][2]==1:
-            labels='Fresh Oranges'
-        elif arr[0][3]==1:
-            labels='Rotten Apples'
-        elif arr[0][4]==1:
-            labels='Rotten Banana'
-        elif arr[0][5]==1:
-            labels='Rotten Oranges'
-
-        return {"result":labels}
-    except Exception as e:
-        traceback.print_exc()
-        response.status_code = 500
-        return str(e)
+import os
+import uvicorn
+import numpy as np
+import traceback
+from io import BytesIO
+
+from pydantic import BaseModel
+from urllib.request import Request
+from fastapi import FastAPI, Response, UploadFile
+from utils import load_image_into_numpy_array
+from tensorflow.keras.models import load_model
+from tensorflow.keras.preprocessing.image import img_to_array
+from tensorflow.keras.preprocessing.image import load_img, img_to_array
+
+app = FastAPI()
+
+@app.post("/predict_image")
+async def predict_image(img: UploadFile, response: Response):
+    try:
+        # Checking if it's an image
+        if img.content_type not in ["image/jpeg", "image/png"]:
+            response.status_code = 400
+            return "File is Not an Image"
+
+        # Read file content
+        file_content = await img.read()
+
+        # Convert bytes to a file-like object
+        file_like_object = BytesIO(file_content)
+
+        # Load the image from the file-like object
+        img = load_img(file_like_object, target_size=(150, 150))
+
+        # Load the model
+        model = load_model('model1.h5')
+
+        # Prepare the image for prediction
+        image = img_to_array(img)
+        image = np.expand_dims(image, axis=0)
+
+        # Predict the class of the image
+        arr = model.predict(image)
+
+        if arr[0][0]==1:
+            labels='Fresh Apples'
+        elif arr[0][1]==1:
+            labels='Fresh Banana'
+        elif arr[0][2]==1:
+            labels='Fresh Cucumber'
+        elif arr[0][3]==1:
+            labels='Fresh Okra'
+        elif arr[0][4]==1:
+            labels='Fresh Oranges'
+        elif arr[0][5]==1:
+            labels='Fresh Potato'
+        elif arr[0][6]==1:
+            labels='Fresh Tomato'
+        elif arr[0][7]==1:
+            labels='Rotten Apples'
+        elif arr[0][8]==1:
+            labels='Rotten Banana'
+        elif arr[0][9]==1:
+            labels='Rotten Cucumber'
+        elif arr[0][10]==1:
+            labels='Rotten Okra'
+        elif arr[0][11]==1:
+            labels='Rotten Oranges'
+        elif arr[0][12]==1:
+            labels='Rotten Potato'
+        elif arr[0][13]==1:
+            labels='Rotten Tomato'
+
+        return {"result":labels}
+    except Exception as e:
+        traceback.print_exc()
+        response.status_code = 500
+        return str(e)

dataset

@@ -0,0 +1,2 @@
+FROM KAGGLE : 
+https://www.kaggle.com/datasets/raghavrpotdar/fresh-and-stale-images-of-fruits-and-vegetables

labels.txt

@@ -0,0 +1,27 @@
+'Fresh Apples'
+
+'Fresh Banana'
+
+'Fresh Cucumber'
+
+'Fresh Okra'
+
+'Fresh Oranges'
+
+'Fresh Potato'
+
+'Fresh Tomato'
+
+'Rotten Apples'
+
+'Rotten Banana'
+
+'Rotten Cucumber'
+
+'Rotten Okra'
+
+'Rotten Oranges'
+
+'Rotten Potato'
+
+'Rotten Tomato'

requirements.txt

@@ -1,9 +1,9 @@
-fastapi
-pydantic
-uvicorn[standard]
-python-multipart
-numpy
-Pillow
-opencv-python
-tensorflow
+fastapi
+pydantic
+uvicorn[standard]
+python-multipart
+numpy
+Pillow
+opencv-python
+tensorflow
 keras

utils.py

@@ -1,7 +1,7 @@
-import numpy as np
-
-from PIL import Image
-from io import BytesIO
-
-def load_image_into_numpy_array(data):
+import numpy as np
+
+from PIL import Image
+from io import BytesIO
+
+def load_image_into_numpy_array(data):
     return np.array(Image.open(BytesIO(data)))