Initial commit

.ipynb_checkpoints/Project-checkpoint.ipynb

+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "21a304bb",
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "IndentationError",
+     "evalue": "expected an indented block after 'for' statement on line 57 (1098499893.py, line 59)",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;36m  Cell \u001b[1;32mIn[10], line 59\u001b[1;36m\u001b[0m\n\u001b[1;33m    model = define_model()\u001b[0m\n\u001b[1;37m    ^\u001b[0m\n\u001b[1;31mIndentationError\u001b[0m\u001b[1;31m:\u001b[0m expected an indented block after 'for' statement on line 57\n"
+     ]
+    }
+   ],
+   "source": [
+    "# baseline cnn model for mnist\n",
+    "from numpy import mean\n",
+    "from numpy import std\n",
+    "from matplotlib import pyplot as plt\n",
+    "from sklearn.model_selection import KFold\n",
+    "from tensorflow.keras.datasets import mnist\n",
+    "from tensorflow.keras.utils import to_categorical\n",
+    "from tensorflow.keras.models import Sequential\n",
+    "from tensorflow.keras.layers import Conv2D\n",
+    "from tensorflow.keras.layers import MaxPooling2D\n",
+    "from tensorflow.keras.layers import Dense\n",
+    "from tensorflow.keras.layers import Flatten\n",
+    "from tensorflow.keras.optimizers import SGD\n",
+    " \n",
+    "# load train and test dataset\n",
+    "def load_dataset():\n",
+    " # load dataset\n",
+    " (trainX, trainY), (testX, testY) = mnist.load_data()\n",
+    " # reshape dataset to have a single channel\n",
+    " trainX = trainX.reshape((trainX.shape[0], 28, 28, 1))\n",
+    " testX = testX.reshape((testX.shape[0], 28, 28, 1))\n",
+    " # one hot encode target values\n",
+    " trainY = to_categorical(trainY)\n",
+    " testY = to_categorical(testY)\n",
+    " return trainX, trainY, testX, testY\n",
+    " \n",
+    "# scale pixels\n",
+    "def prep_pixels(train, test):\n",
+    " # convert from integers to floats\n",
+    " train_norm = train.astype('float32')\n",
+    " test_norm = test.astype('float32')\n",
+    " # normalize to range 0-1\n",
+    " train_norm = train_norm / 255.0\n",
+    " test_norm = test_norm / 255.0\n",
+    " # return normalized images\n",
+    " return train_norm, test_norm\n",
+    " \n",
+    "# define cnn model\n",
+    "def define_model():\n",
+    " model = Sequential()\n",
+    " model.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', input_shape=(28, 28, 1)))\n",
+    " model.add(MaxPooling2D((2, 2)))\n",
+    " model.add(Flatten())\n",
+    " model.add(Dense(100, activation='relu', kernel_initializer='he_uniform'))\n",
+    " model.add(Dense(10, activation='softmax'))\n",
+    " # compile model\n",
+    " opt = SGD(learning_rate=0.01, momentum=0.9)\n",
+    " model.compile(optimizer=opt, loss='categorical_crossentropy', metrics=['accuracy'])\n",
+    " return model\n",
+    " \n",
+    "# evaluate a model using k-fold cross-validation\n",
+    "def evaluate_model(dataX, dataY, n_folds=5):\n",
+    " scores, histories = list(), list()\n",
+    " # prepare cross validation\n",
+    " kfold = KFold(n_folds, shuffle=True, random_state=1)\n",
+    " # enumerate splits\n",
+    " for train_ix, test_ix in kfold.split(dataX):\n",
+    " # define model\n",
+    " model = define_model()\n",
+    " # select rows for train and test\n",
+    " trainX, trainY, testX, testY = dataX[train_ix], dataY[train_ix], dataX[test_ix], dataY[test_ix]\n",
+    " # fit model\n",
+    " history = model.fit(trainX, trainY, epochs=10, batch_size=32, validation_data=(testX, testY), verbose=0)\n",
+    " # evaluate model\n",
+    " _, acc = model.evaluate(testX, testY, verbose=0)\n",
+    " print('> %.3f' % (acc * 100.0))\n",
+    " # stores scores\n",
+    " scores.append(acc)\n",
+    " histories.append(history)\n",
+    " return scores, histories\n",
+    " \n",
+    "# plot diagnostic learning curves\n",
+    "def summarize_diagnostics(histories):\n",
+    " for i in range(len(histories)):\n",
+    " # plot loss\n",
+    " plt.subplot(2, 1, 1)\n",
+    " plt.title('Cross Entropy Loss')\n",
+    " plt.plot(histories[i].history['loss'], color='blue', label='train')\n",
+    " plt.plot(histories[i].history['val_loss'], color='orange', label='test')\n",
+    " # plot accuracy\n",
+    " plt.subplot(2, 1, 2)\n",
+    " plt.title('Classification Accuracy')\n",
+    " plt.plot(histories[i].history['accuracy'], color='blue', label='train')\n",
+    " plt.plot(histories[i].history['val_accuracy'], color='orange', label='test')\n",
+    " plt.show()\n",
+    " \n",
+    "# summarize model performance\n",
+    "def summarize_performance(scores):\n",
+    " # print summary\n",
+    " print('Accuracy: mean=%.3f std=%.3f, n=%d' % (mean(scores)*100, std(scores)*100, len(scores)))\n",
+    " # box and whisker plots of results\n",
+    " plt.boxplot(scores)\n",
+    " plt.show()\n",
+    " \n",
+    "# run the test harness for evaluating a model\n",
+    "def run_test_harness():\n",
+    " # load dataset\n",
+    " trainX, trainY, testX, testY = load_dataset()\n",
+    " # prepare pixel data\n",
+    " trainX, testX = prep_pixels(trainX, testX)\n",
+    " # evaluate model\n",
+    " scores, histories = evaluate_model(trainX, trainY)\n",
+    " # learning curves\n",
+    " summarize_diagnostics(histories)\n",
+    " # summarize estimated performance\n",
+    " summarize_performance(scores)\n",
+    " \n",
+    "# entry point, run the test harness\n",
+    "run_test_harness()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "38771ec7",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

main.py

+import os
+import PIL
+import cv2
+import glob
+import numpy as np
+from tkinter import *
+from PIL import Image, ImageDraw, ImageGrab
+
+# Tkinter done using https://medium.com/analytics-vidhya/handwritten-digit-recognition-gui-app-46e3d7b37287
+
+def Clear_Widget():
+    global cv
+    cv.delete("all")
+
+def Activate_Event(event):
+    global lastx, lasty
+    cv.bind('<B1-Motion>', Draw_Lines)
+    lastx, lasty = event.x, event.y
+
+def Draw_Lines(event):
+    global lastx, lasty
+    x, y = event.x, event.y
+    cv.create_line((lastx, lasty, x, y), width=8, fill='black', capstyle=ROUND, smooth=TRUE, splinesteps=12)
+    lastx, lasty = x, y
+
+def Recognise_Digit():
+    global image_number
+    predictions = []
+    percentage = []
+    filename = f'image_{image_number}.png'
+    widget = cv
+
+    x=root.winfo_rootx()+widget.winfo_x()
+    y=root.winfo_rooty()+widget.winfo_y()
+    x1=x+widget.winfo_width()
+    y1=y+widget.winfo_height()
+
+    ImageGrab.grab().crop((x,y,x1,y1)).save(filename)
+
+    image = cv2.imread(filename, cv2.IMREAD_COLOR)
+    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
+    ret,th = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)
+    contours = cv2.findContours(th, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[0]
+
+    for cnt in contours:
+        x,y,w,h = cv2.boundingRect(cnt)
+        cv2.rectangle(image, (x,y), (x+w, y+h), (255,0,0), 1)
+        top = int(0.05 * th.shape[0])
+        bottom = top
+        left = int(0.05 * th.shape[1])
+        right = left
+        th_up = cv2.copyMakeBorder(th, top, bottom, left, right, cv2.BORDER_REPLICATE)
+        roi = th[y-top:y+h+bottom, x-left:x+w+right]
+        img = cv2.resize(roi, (28, 28), interpolation=cv2.INTER_AREA)
+        img = img.reshape(1,28,28,1)
+        img = img/255.0
+        pred = model.predict([img])[0]
+        final_pred = np.argmax(pred)
+        data = str(final_pred) + ' ' + str(int(max(pred)*100))+'%'
+        font = cv2.FONT_HERSHEY_SIMPLEX
+        fontScale = 0.5
+        color = (255,0,0)
+        thickness = 1
+        cv2.putText(image, data, (x,y-5), font, fontScale, color, thickness)
+
+    cv2.imshow('image', image)
+    cv2.waitKey(0)
+
+# Load model
+from keras.models import load_model
+model = load_model(r'D:\University\AI\Project\model.h5')
+
+# Create a main window
+root = Tk()
+root.resizable(0, 0)
+root.title("Digit Recognition")
+
+# Initialise variables
+lastx, lasty = None, None
+image_number = 0
+
+# Create canvas
+cv = Canvas(root, width=640, height=480, bg='white')
+cv.grid(row=0, column=0, pady=2, sticky=W, columnspan=2)
+
+# Deal with events
+cv.bind('<Button-1>', Activate_Event)
+
+# Add buttons
+btn_save = Button(text="Recognise Digit", command=Recognise_Digit)
+btn_save.grid(row=2, column=0, pady=1, padx=1)
+btn_clear = Button(text="Clear", command=Clear_Widget)
+btn_clear.grid(row=2, column=1, pady=1, padx=1)
+
+root.mainloop()
\ No newline at end of file