noshot 12.0.0__py3-none-any.whl → 14.0.0__py3-none-any.whl

noshot/data/DLE FSD BDA/DLE/DLE 4 (senior)/Exp03/DL-Ex3-RNN.ipynb

@@ -0,0 +1,591 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Construct an object detector using Convolutional Neural Networks"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os,cv2,keras\n",
+    "import pandas as pd\n",
+    "import matplotlib.pyplot as plt\n",
+    "import numpy as np\n",
+    "import tensorflow as tf"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import cv2\n",
+    "print(cv2.__version__)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "path = \"C://Users//admin//DataSet RCNN//DataSet RCNN//Images//Images\"\n",
+    "annot = \"C://Users//admin//DataSet RCNN//DataSet RCNN//Airplanes_Annotations//Airplanes_Annotations\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for e,i in enumerate(os.listdir(annot)):\n",
+    "    if e < 10:\n",
+    "        filename = i.split(\".\")[0]+\".jpg\"\n",
+    "        print(filename)\n",
+    "        img = cv2.imread(os.path.join(path,filename))\n",
+    "        df = pd.read_csv(os.path.join(annot,i))\n",
+    "        plt.imshow(img)\n",
+    "        for row in df.iterrows():\n",
+    "            x1 = int(row[1][0].split(\" \")[0])\n",
+    "            y1 = int(row[1][0].split(\" \")[1])\n",
+    "            x2 = int(row[1][0].split(\" \")[2])\n",
+    "            y2 = int(row[1][0].split(\" \")[3])\n",
+    "            cv2.rectangle(img,(x1,y1),(x2,y2),(255,0,0), 2)\n",
+    "        plt.figure()\n",
+    "        plt.imshow(img)\n",
+    "        break"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#!pip uninstall opencv-contrib-python opencv-python"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#!pip install opencv-contrib-python --user"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from cv2.ximgproc import segmentation\n",
+    "#from cv2.ximgproc.segmentation import createSelectiveSearchSegmentation\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import cv2\n",
+    "import cv2.ximgproc.segmentation\n",
+    "\n",
+    "print(dir(cv2.ximgproc.segmentation))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from cv2.ximgproc.segmentation import createSelectiveSearchSegmentation"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "cv2.setUseOptimized(True);\n",
+    "#ss = segmentation.createSelectiveSearchSegmentation()\n",
+    "ss = createSelectiveSearchSegmentation()\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "im = cv2.imread(os.path.join(path,\"airplane_687.jpg\"))\n",
+    "ss.setBaseImage(im)\n",
+    "ss.switchToSelectiveSearchFast()\n",
+    "rects = ss.process()\n",
+    "imOut = im.copy()\n",
+    "for i, rect in (enumerate(rects)):\n",
+    "    x, y, w, h = rect\n",
+    "#     print(x,y,w,h)\n",
+    "#     imOut = imOut[x:x+w,y:y+h]\n",
+    "    cv2.rectangle(imOut, (x, y), (x+w, y+h), (0, 255, 0), 1, cv2.LINE_AA)\n",
+    "# plt.figure()\n",
+    "plt.imshow(imOut)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "train_images=[]\n",
+    "train_labels=[]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def get_iou(bb1, bb2):\n",
+    "    assert bb1['x1'] < bb1['x2']\n",
+    "    assert bb1['y1'] < bb1['y2']\n",
+    "    assert bb2['x1'] < bb2['x2']\n",
+    "    assert bb2['y1'] < bb2['y2']\n",
+    "\n",
+    "    x_left = max(bb1['x1'], bb2['x1'])\n",
+    "    y_top = max(bb1['y1'], bb2['y1'])\n",
+    "    x_right = min(bb1['x2'], bb2['x2'])\n",
+    "    y_bottom = min(bb1['y2'], bb2['y2'])\n",
+    "\n",
+    "    if x_right < x_left or y_bottom < y_top:\n",
+    "        return 0.0\n",
+    "\n",
+    "    intersection_area = (x_right - x_left) * (y_bottom - y_top)\n",
+    "\n",
+    "    bb1_area = (bb1['x2'] - bb1['x1']) * (bb1['y2'] - bb1['y1'])\n",
+    "    bb2_area = (bb2['x2'] - bb2['x1']) * (bb2['y2'] - bb2['y1'])\n",
+    "\n",
+    "    iou = intersection_area / float(bb1_area + bb2_area - intersection_area)\n",
+    "    assert iou >= 0.0\n",
+    "    assert iou <= 1.0\n",
+    "    return iou"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "ss = cv2.ximgproc.segmentation.createSelectiveSearchSegmentation()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for e,i in enumerate(os.listdir(annot)):\n",
+    "    try:\n",
+    "        if i.startswith(\"airplane\"):\n",
+    "            filename = i.split(\".\")[0]+\".jpg\"\n",
+    "            print(e,filename)\n",
+    "            image = cv2.imread(os.path.join(path,filename))\n",
+    "            df = pd.read_csv(os.path.join(annot,i))\n",
+    "            gtvalues=[]\n",
+    "            for row in df.iterrows():\n",
+    "                x1 = int(row[1][0].split(\" \")[0])\n",
+    "                y1 = int(row[1][0].split(\" \")[1])\n",
+    "                x2 = int(row[1][0].split(\" \")[2])\n",
+    "                y2 = int(row[1][0].split(\" \")[3])\n",
+    "                gtvalues.append({\"x1\":x1,\"x2\":x2,\"y1\":y1,\"y2\":y2})\n",
+    "            ss.setBaseImage(image)\n",
+    "            ss.switchToSelectiveSearchFast()\n",
+    "            ssresults = ss.process()\n",
+    "            imout = image.copy()\n",
+    "            counter = 0\n",
+    "            falsecounter = 0\n",
+    "            flag = 0\n",
+    "            fflag = 0\n",
+    "            bflag = 0\n",
+    "            for e,result in enumerate(ssresults):\n",
+    "                if e < 200 and flag == 0:\n",
+    "                    for gtval in gtvalues:\n",
+    "                        x,y,w,h = result\n",
+    "                        iou = get_iou(gtval,{\"x1\":x,\"x2\":x+w,\"y1\":y,\"y2\":y+h})\n",
+    "                        if counter < 30:\n",
+    "                            if iou > 0.70:\n",
+    "                                timage = imout[y:y+h,x:x+w]\n",
+    "                                resized = cv2.resize(timage, (56,56), interpolation = cv2.INTER_AREA)\n",
+    "                                train_images.append(resized)\n",
+    "                                train_labels.append(1)\n",
+    "                                counter += 1\n",
+    "                        else :\n",
+    "                            fflag =1\n",
+    "                        if falsecounter <30:\n",
+    "                            if iou < 0.3:\n",
+    "                                timage = imout[y:y+h,x:x+w]\n",
+    "                                resized = cv2.resize(timage, (56,56), interpolation = cv2.INTER_AREA)\n",
+    "                                train_images.append(resized)\n",
+    "                                train_labels.append(0)\n",
+    "                                falsecounter += 1\n",
+    "                        else :\n",
+    "                            bflag = 1\n",
+    "                    if fflag == 1 and bflag == 1:\n",
+    "                        print(\"inside\")\n",
+    "                        flag = 1\n",
+    "    except Exception as e:\n",
+    "        print(e)\n",
+    "        print(\"error in \"+filename)\n",
+    "        continue"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "X_new = np.array(train_images)\n",
+    "y_new = np.array(train_labels)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "X_new.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from keras.layers import Dense\n",
+    "from keras import Model\n",
+    "from keras import optimizers\n",
+    "from keras.preprocessing.image import ImageDataGenerator\n",
+    "from keras.applications.vgg16 import VGG16"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "vggmodel = VGG16(weights='imagenet', include_top=True)\n",
+    "vggmodel.summary()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import tensorflow as tf\n",
+    "from tensorflow.keras import layers, models\n",
+    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
+    "import matplotlib.pyplot as plt\n",
+    "import seaborn as sns\n",
+    "import numpy as np\n",
+    "from sklearn.metrics import classification_report, confusion_matrix"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model = models.Sequential([\n",
+    "    layers.Conv2D(32, (3, 3), activation='relu', input_shape=(56, 56, 3)),\n",
+    "    layers.MaxPooling2D((2, 2)),\n",
+    "    layers.Conv2D(64, (3, 3), activation='relu'),\n",
+    "    layers.MaxPooling2D((2, 2)),\n",
+    "    layers.Conv2D(128, (3, 3), activation='relu'),\n",
+    "    layers.MaxPooling2D((2, 2)),\n",
+    "    layers.Conv2D(128, (3, 3), activation='relu'),\n",
+    "    layers.MaxPooling2D((2, 2)),\n",
+    "    layers.Flatten(),\n",
+    "    layers.Dense(512, activation='relu'),\n",
+    "])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# for layers in (vggmodel.layers)[:15]:\n",
+    "#     print(layers)\n",
+    "#     layers.trainable = False"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "X= model.layers[-2].output"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "predictions = Dense(2, activation=\"softmax\")(X)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model_final = Model(inputs = model.input, outputs = predictions)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from keras.optimizers import Adam\n",
+    "opt = Adam(learning_rate=0.0001)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model_final.compile(loss = keras.losses.categorical_crossentropy, optimizer = opt, metrics=[\"accuracy\"])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model_final.summary()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.model_selection import train_test_split\n",
+    "from sklearn.preprocessing import LabelBinarizer"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "class MyLabelBinarizer(LabelBinarizer):\n",
+    "    def transform(self, y):\n",
+    "        Y = super().transform(y)\n",
+    "        if self.y_type_ == 'binary':\n",
+    "            return np.hstack((Y, 1-Y))\n",
+    "        else:\n",
+    "            return Y\n",
+    "    def inverse_transform(self, Y, threshold=None):\n",
+    "        if self.y_type_ == 'binary':\n",
+    "            return super().inverse_transform(Y[:, 0], threshold)\n",
+    "        else:\n",
+    "            return super().inverse_transform(Y, threshold)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "lenc = MyLabelBinarizer()\n",
+    "Y =  lenc.fit_transform(y_new)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "X_train, X_test , y_train, y_test = train_test_split(X_new,Y,test_size=0.10)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print(X_train.shape,X_test.shape,y_train.shape,y_test.shape)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "trdata = ImageDataGenerator(horizontal_flip=True, vertical_flip=True, rotation_range=90)\n",
+    "traindata = trdata.flow(x=X_train, y=y_train)\n",
+    "tsdata = ImageDataGenerator(horizontal_flip=True, vertical_flip=True, rotation_range=90)\n",
+    "testdata = tsdata.flow(x=X_test, y=y_test)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from keras.callbacks import ModelCheckpoint, EarlyStopping"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "checkpoint = ModelCheckpoint(\"ieeercnn_vgg16_1.h5\", monitor='val_loss', verbose=1, save_best_only=True, save_weights_only=False, mode='auto', period=1)\n",
+    "early = EarlyStopping(monitor='val_loss', min_delta=0, patience=100, verbose=1, mode='auto')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "hist = model_final.fit_generator(generator= traindata, steps_per_epoch= 10, epochs= 1000, validation_data= testdata, validation_steps=2, callbacks=[checkpoint,early])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import matplotlib.pyplot as plt\n",
+    "# plt.plot(hist.history[\"acc\"])\n",
+    "# plt.plot(hist.history['val_acc'])\n",
+    "plt.plot(hist.history['loss'])\n",
+    "plt.plot(hist.history['val_loss'])\n",
+    "plt.title(\"model loss\")\n",
+    "plt.ylabel(\"Loss\")\n",
+    "plt.xlabel(\"Epoch\")\n",
+    "plt.legend([\"Loss\",\"Validation Loss\"])\n",
+    "plt.show()\n",
+    "plt.savefig('chart loss.png')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "im = X_test[200]\n",
+    "plt.imshow(im)\n",
+    "img = np.expand_dims(im, axis=0)\n",
+    "out= model_final.predict(img)\n",
+    "if out[0][0] > out[0][1]:\n",
+    "    print(\"not plane\")\n",
+    "else:\n",
+    "    print(\"plane\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "z = 0\n",
+    "\n",
+    "for e, i in enumerate(os.listdir(path)):\n",
+    "    if i.startswith(\"airplane_007\"):\n",
+    "        z += 1\n",
+    "        img = cv2.imread(os.path.join(path, i))\n",
+    "        ss.setBaseImage(img)\n",
+    "        ss.switchToSelectiveSearchFast()\n",
+    "        ssresults = ss.process()\n",
+    "        imout = img.copy()\n",
+    "\n",
+    "        for e, result in enumerate(ssresults):\n",
+    "            if e < 2000:\n",
+    "                x, y, w, h = result\n",
+    "                timage = imout[y:y+h, x:x+w]\n",
+    "                resized = cv2.resize(timage, (56, 56), interpolation=cv2.INTER_AREA)  # Resize to (56, 56)\n",
+    "                img = np.expand_dims(resized, axis=0)\n",
+    "                \n",
+    "                # Ensure input shape matches the model's expected input shape\n",
+    "                out = model_final.predict(img)\n",
+    "                \n",
+    "                if out[0][0] > 0.65:\n",
+    "                    cv2.rectangle(imout, (x, y), (x+w, y+h), (0, 255, 0), 1, cv2.LINE_AA)\n",
+    "\n",
+    "        plt.figure()\n",
+    "        plt.imshow(cv2.cvtColor(imout, cv2.COLOR_BGR2RGB))\n",
+    "        plt.show()"
+   ]
+  }
+ ],
+ "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": "3.12.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}