From 063775050550fde7a100998723f2b466b66add92 Mon Sep 17 00:00:00 2001
From: Vignesh skanda <agvskanda@gmail.com>
Date: Sun, 22 Sep 2024 13:10:23 +0530
Subject: [PATCH] Update AI.py

Updated the code as it needed few changes in it.
---
 AI/AI.py | 66 ++++++++++++++++++++------------------------------------
 1 file changed, 23 insertions(+), 43 deletions(-)

diff --git a/AI/AI.py b/AI/AI.py
index c388774..8a809e7 100644
--- a/AI/AI.py
+++ b/AI/AI.py
@@ -3,50 +3,30 @@
 import os
 from tensorflow import keras
 
-
-#(x_train, y_train), (x_test, y_test) = keras.datasets.mnist.load_data()
-
-## modify the training data
-#x_train = x_train.reshape((-1, 28, 28, 1))
-#x_train = x_train / 255.0
-
-## build the neural network model
-#model = keras.Sequential([
-#    keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)),
-#    keras.layers.MaxPooling2D((2, 2)),
-#    keras.layers.Flatten(),
-#    keras.layers.Dense(10, activation='softmax')
-#])
-
-## compile the model
-#model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
-
-## train the model
-#model.fit(x_train, y_train, epochs=5, validation_data=(x_test, y_test))
-
-## save the model to a file
-#model.save('model.h5')
-
-
-# load the model from a file
-model = keras.models.load_model('model.h5')
-
+# Load the model from a file
+try:
+    model = keras.models.load_model('model.h5')
+except FileNotFoundError:
+    print("Model file not found!")
+    exit(1)
+
+# Define a function to load and preprocess an image
+def load_image(file_name: str) -> np.ndarray:
+    img = Image.open(f'img/{file_name}').convert('L')
+    img = img.resize((28, 28))
+    img = np.array(img) / 255.0
+    img = img.reshape((-1, 28, 28, 1))
+    return img
+
+# Iterate over the images in the directory
 for file_name in os.listdir('img'):
     if file_name.endswith(".png"):
-        # load and convert the image to a numpy array
-        img = Image.open(f'img/{file_name}').convert('L')
-        img = img.resize((28, 28))
-        img = np.array(img)
-
-        # normalize the image data
-        img = img / 255.0
-
-        # reshape the image to a shape accepted by the model
-        img = img.reshape((-1, 28, 28, 1))
-
-        # make a prediction on the image
-        pred = model.predict(img)
-        digit = np.argmax(pred)
+        try:
+            img = load_image(file_name)
+            pred = model.predict(img)
+            digit = np.argmax(pred)
+            print(f"Recognized digit for file {file_name} is: {digit}")
+        except Exception as e:
+            print(f"Error processing file {file_name}: {e}")
 
-        print(f"Recognized digit for file {file_name} is: {digit}")
 input("Press Enter to end...")