Day-7_SUMMER_TRAINING_AI/ML IMAGE REC. 🤔

day_7_dhruvdhayal_ai_ml (1).py

@@ -0,0 +1,272 @@
+# -*- coding: utf-8 -*-
+"""Day-7_DHRUVDHAYAL_AI/ML.ipynb
+
+Automatically generated by Colab.
+
+Original file is located at
+    https://colab.research.google.com/drive/1K26xv32kLaOTY8qLVpCVxFylXPeMLQx7
+
+#Now, we directly acess the data values from the Drive Directly!
+"""
+
+from google.colab import drive;
+drive.mount('/gdrive');
+
+!unzip '/content/drive/MyDrive/Colab Notebooks/orl_face/orl_face.zip' -d '/content/drive/MyDrive/Colab Notebooks/orl_face'
+
+#Now, Importing the Images data from the unzip files.
+import matplotlib.pyplot as plt;
+import matplotlib.image as mimg;
+import numpy as np;
+import pandas as pd;
+
+#Now, we need to read the values of the Data from a Drive.
+user_name=24; #Labels.
+samples_no=6; #Define the Variable Sample_Numbers.
+path='/content/drive/MyDrive/Colab Notebooks/orl_face/orl_face/u%d/%d.png'%(user_name,samples_no);
+
+#Now, we need to read the data from images.
+imag=mimg.imread(path);
+print("\n 1. Type of the Image is: ",type(imag));
+print("\n 2. Length of the Image is: ",imag.shape);
+print("\n");
+
+#Now, we need to plot the Image.
+plt.figure(1,figsize=(5,10));
+plt.imshow(imag,cmap='gray');
+plt.axis("off");
+plt.show();
+
+#Now, convert 2-D Images it into 1-D Images by using the Flatten.
+feat=imag.reshape(1,-1);
+print("\n 1. Length of the Images: ",imag.shape);
+print("\n 2. Length of the Features: ",feat.shape);
+print("\n --> Range: ",imag.min()," - ",imag.max());
+
+#Logic to acess all the samples of the User.
+#Of the Valued Users.
+total_sample=400;
+user_name=26;
+sample_no=6;
+data=np.zeros((total_sample,imag.shape[0]*imag.shape[1]));
+label=np.zeros((total_sample));
+images=np.zeros((total_sample,imag.shape[0],imag.shape[1]));
+index=-1;
+for i in range(1,41,1):
+  for j in range(1,11,1):
+    index=index+1;
+    #acess any of the single image.
+    user_name=i;
+    sample_no=j;
+    path='/content/drive/MyDrive/Colab Notebooks/orl_face/orl_face/u%d/%d.png'%(user_name,sample_no);
+    #reading the Image.
+    im=mimg.imread(path);
+    data[index,:]=feat
+    label[index]=i
+    images[index,:,:]=im
+    print("user num ",i,'samp no',j,'processed...');
+
+#Now, displaying the values of the Image.
+plt.imshow(images[397,:,:],cmap='gray');
+plt.axis('off');
+plt.show();
+
+"""#Now, we need to represent the values of the (5*5) Matrix."""
+
+#Importing all the Built-in Libraries.
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt # Import matplotlib.pyplot
+
+#Now, we need to show all the (5*5) Matrix Values.
+plt.figure(1,figsize=(10,10));
+for i in range(5):
+  for j in range(5):
+    num=np.random.randint(0,400);
+    samples=images[num,:,:];
+    la=label[num];
+    plt.subplot(5,5,i*5+j+1) # Use plt to access subplot
+    plt.imshow(samples,cmap='gray');
+    plt.axis('off');
+    plt.title(" "+str(int(la)));
+#showing the Image (5*5) Matrix.
+plt.show();
+
+from sklearn import svm
+import pandas as pd
+X = data # Extract the data from the 'data' key
+y =label # Use the target variable from the Iris dataset
+
+
+# split the data into 70:30 ratio
+Xtrain,Xtest,ytrain,ytest = model_selection.train_test_split(X,y,test_size=0.3,random_state=5)
+print(Xtrain.shape,ytrain.shape)
+print(Xtest.shape,ytest.shape)
+
+ker = ['poly','linear','rbf']
+c_value = [1,2,3]
+
+# pre allocation of the result variable
+result = np.zeros((len(ker),len(c_value)))
+for i in range(len(ker)):
+  for j in range(len(c_value)):
+    # create the svm classifier
+    orl_svm_model = svm.SVC(kernel=ker[i],gamma='scale',C=c_value[j])
+
+    # train the model
+    orl_svm_model = orl_svm_model.fit(Xtrain,ytrain)
+
+    # predict the labels
+    ypred = orl_svm_model.predict(Xtest)
+
+    # accuracy
+    acc = metrics.accuracy_score(ypred,ytest)
+    #print("accuracy:", acc)
+    result[i,j]=acc
+print(result)
+
+ResultDF = pd.DataFrame(result,index=ker,columns=["C=1","C=2","C=3"])
+ResultDF
+print("\n Showing in the form of the Graphical Methods!");
+ResultDF.plot(kind='bar',figsize=(4,4))
+
+#Saving our trained model in the Google Drive.
+#We, use the Joblib to save the models.
+import joblib
+# final best model
+# kernel function - linear , C =1
+
+orl_svm_model = svm.SVC(kernel='linear',gamma='scale',C=1)
+# train the model
+orl_svm_model = orl_svm_model.fit(Xtrain,ytrain)
+
+# save the trained model
+joblib.dump(orl_svm_model,'/content/drive/MyDrive/Colab Notebooks/orl_face_model.pkl')
+
+#Show the random image from 1-10 , asked by the user to take as input form of parameter.
+#Ask the user as ID to show that location image randomly.
+import matplotlib.pyplot as plt;
+import matplotlib.image as mimg;
+import pandas as pd;
+import numpy as np;
+
+#Now, we simply ask from the User.
+user_name=int(input("Enter the User ID: "));
+sample_no=np.random.randint(8,10);
+path='/content/drive/MyDrive/Colab Notebooks/orl_face/orl_face/u%d/%d.png'%(user_name,sample_no);
+
+#Now, we need to show the value of the Image.
+imag=mimg.imread(path);
+print("\n");
+plt.figure(1,figsize=(5,10));
+plt.imshow(imag,cmap='gray');
+plt.axis("off");
+plt.title('Query Image');
+plt.show();
+
+#We, make the trained model and save in the drive by using the joblib.
+from sklearn import datasets, svm, metrics, model_selection;
+import joblib;
+import numpy as np;
+import pandas as pd;
+
+feat=imag.reshape(1,-1);
+
+#Set the value of the Path_model where, we have to store the model in pkl. file on Drive.
+path_model='/content/drive/MyDrive/Colab Notebooks/orl_face_model.pkl';
+
+#Train the Model by the help of Classifications.
+face_model=joblib.load(path_model);
+
+#predict the image model by datasets.
+result=face_model.predict(feat);
+
+print("Prediction: ",int(result[0]));
+
+#set the location of the zip image file.
+path='/content/drive/MyDrive/Colab Notebooks/orl_face/orl_face/u%d/%d.png'%(result[0],1);
+
+im_pred = mimg.imread(path)
+
+plt.figure(1,(8,5))
+plt.subplot(1,2,1)
+plt.imshow(im,cmap='gray')
+plt.axis('off')
+plt.title("Query image")
+
+plt.subplot(1,2,2)
+plt.imshow(im_pred,cmap='gray')
+plt.axis('off')
+plt.title("predicted user")
+
+"""#Example of FP-(FALSE - POSITIVE), Very, Dangerous I check but I don't know where is the problem where all snippet is found to be correct, I refer to my friend's colab notebook for reviewing."""
+
+import matplotlib.image as mimg
+import matplotlib.pyplot as plt
+import numpy as np
+
+# load the query image
+usrId = int(input("Enter the user number:"))
+samp = np.random.randint(1,10) # selecting any random sample number
+path = "/content/drive/MyDrive/Colab Notebooks/orl_face/orl_face/u%d/%d.png"%(usrId,samp)
+
+im = mimg.imread(path)
+
+plt.figure(1,(5,5))
+plt.imshow(im,cmap='gray')
+plt.axis('off')
+plt.title("Query image")
+
+import joblib
+from sklearn import svm ,metrics
+# calcualte the features of the query image
+feat_query = im.reshape(1,-1)
+
+# model path
+path_model = '/content/drive/MyDrive/Colab Notebooks/orl_face_model.pkl';
+# load the trained model
+face_model = joblib.load(path_model)
+
+
+# predict the id of the query image
+id = face_model.predict(feat_query)
+print("User id predicted is :",int(id[0]))
+
+
+path = "/content/drive/MyDrive/Colab Notebooks/orl_face/orl_face/u%d/%d.png"%(id[0],1)
+
+im_pred = mimg.imread(path)
+
+plt.figure(1,(8,5))
+plt.subplot(1,2,1)
+plt.imshow(im,cmap='gray')
+plt.axis('off')
+plt.title("Query image")
+
+plt.subplot(1,2,2)
+plt.imshow(im_pred,cmap='gray')
+plt.axis('off')
+plt.title("predicted user")
+
+"""#False-Positive!
+
+#Now, working on Generating an Computer Applications to analyze the Brain Tumor Problem.
+"""
+
+#Now, unzip the tumor files.
+!unzip '/content/drive/MyDrive/Colab Notebooks/orl_face/new_train-20240708T093236Z-001.zip' -d '/content/drive/MyDrive/Colab Notebooks'
+
+#Finding the Tumor in the Brain with the White-Spot.
+import matplotlib.pyplot as plt;
+import matplotlib.image as mimg;
+import numpy as np;
+plt.figure(1,figsize=(10,10));
+for i in range(1,62,1):
+  path='/content/drive/MyDrive/Colab Notebooks/new_train/d(%d).png'%(i);
+  imag=mimg.imread(path);
+  plt.subplot(8,8,i);
+  plt.imshow(imag,cmap='gray');
+  plt.title(i);
+  plt.tight_layout();
+  plt.axis("off");
+  #plt.show();