#-----AI2.py Supervised Machine Learning. #--Images are catalogrized into known classes #--Use tensor-flow to build a model for future prediction import cv2 import random import numpy as np import matplotlib.pyplot as plt import os import math import time def cw(k): cv2.waitKey(k) def sh0(w,img): cv2.imshow(w,img) seed = 7 np.random.seed(seed) isx=64; isy=64 #------------------------------------------------------ def Sload_images_from_folder(iszx,iszy,folder): nitems=os.listdir(folder); nsitems=sorted(nitems) images=[]; img5=[] for filename in nsitems: img = cv2.imread(os.path.join(folder,filename), cv2.IMREAD_GRAYSCALE) if img is not None: img1=cv2.resize(img,(iszx,iszy))/255 images.append(img1); img5.append(img) lens=len(nsitems) return lens,nsitems,np.array(images[:lens]),img5 ################################################## def load_images_from_folder(folder,nfile): images = [] for filename in os.listdir(folder): img = cv2.imread(os.path.join(folder,filename), cv2.IMREAD_GRAYSCALE) if img is not None: img = cv2.resize(img,(isx,isy))/255 images.append(img) return np.array(images[:nfile]) ################################################## def AITEST(pathT): # read images #test = load_images_from_folder(pathT,ntest) ntest,fimgs,test,img5=Sload_images_from_folder(isx,isy,pathT) print('load is done path=',pathT) print('fimgs=',fimgs) X_test = test X_test = X_test.reshape(X_test.shape[0], isy,isx, 1).astype("float32") print('X_test:type,len,shape=',type(X_test),len(X_test),X_test.shape) pred = np.argmax(model.predict(X_test),axis=1) #print() #print(X_test) print('pred=',pred) if(1==2): for i in range(ntest): img = X_test[i,:,:,0] print(i,'pred[i]=',pred[i],' LAB=',LAB[pred[i]]) cv2.imshow('img',img) cv2.waitKey(0) ################################################## def buildmodel(X_train,Y_train): # One-hot code Y_train = to_categorical(Y_train); print('Y_train.shape2=',Y_train.shape) # Transform images into 4D tensors X_train = X_train.reshape(X_train.shape[0], isy,isx, 1).astype("float32") X_train,X_test,Y_train,Y_test=train_test_split(X_train,Y_train, \ test_size=0.1, random_state=42) # define the model model = Sequential() model.add(Conv2D(16,(3,3),activation = 'relu', input_shape = (isy,isx,1))) model.add(MaxPooling2D(pool_size=(2,2))) model.add(Conv2D(32,(3,3),activation = 'relu')) model.add(MaxPooling2D(pool_size=(2,2))) model.add(Flatten()) model.add(Dense(64,activation='relu')) model.add(Dropout(0.2)) model.add(Dense(32,activation='relu')) model.add(Dense(ncat,activation='softmax')) model.summary() model.compile(loss = 'categorical_crossentropy', optimizer = "adam", metrics = ['accuracy']) #print(X_train.shape); print(Y_train.shape) history = model.fit(X_train,Y_train,shuffle=True, validation_split=0.1, batch_size=128, epochs=60) # training and validation accuracy acc = history.history["accuracy"] epochs = range(1, len(acc)+1) val_acc = history.history["val_accuracy"] loss = history.history["loss"] epochs = range(1, len(loss)+1) val_loss = history.history["val_loss"] if(1==2): plt.plot(epochs, acc, "bo-", label="Training Acc") plt.plot(epochs, val_acc, "ro--", label="Validation Acc") plt.title("Training and Validation Accuracy") plt.xlabel("Epochs") plt.ylabel("Accuracy") plt.legend() plt.show() # -----------loss------------- plt.plot(epochs, loss, "bo-", label="Training Loss") plt.plot(epochs, val_loss, "ro--", label="Validation Loss") plt.title("Training and Validation Loss") plt.xlabel("Epochs") plt.ylabel("Loss") plt.legend() plt.show() print("\nTesting ...") loss, accuracy = model.evaluate(X_train, Y_train, verbose=0) print("accuracy of training data={:.2f}".format(accuracy)) loss, accuracy = model.evaluate(X_test, Y_test, verbose=0) print("accuracy of testing data={:.2f}".format(accuracy)) model.save(mod1) print('model ',mod1,' is saved...') # set no. of images for each lable KBU=1; path='PNG2468'; ncat=4; nfile=60; ntest=20; pathT=path+'/TEST/' mod1='model_2468_60A.h5' LAB=['2','4','6','8'] print('load path=',path) train = load_images_from_folder(path+'/'+LAB[0]+'/',nfile) print('type(train)=',type(train),train.shape) Y_train = np.zeros(train.shape[0]) X_train = train train = load_images_from_folder(path+'/'+LAB[1]+'/',nfile) Y_train = np.append(Y_train, np.ones(train.shape[0])*1) X_train = np.append(X_train, train, axis=0) #if(ncat==33): # train = load_images_from_folder(path+'/'+LAB[2]+'/',nfile) # Y_train = np.append(Y_train, np.ones(train.shape[0])*2) # X_train = np.append(X_train, train, axis=0) if(ncat>2): for nc in range(2,ncat): train = load_images_from_folder(path+'/'+LAB[nc]+'/',nfile) Y_train = np.append(Y_train, np.ones(train.shape[0])*nc) X_train = np.append(X_train, train, axis=0) print('X_train.shape=',X_train.shape); print('y_train.shape=',Y_train.shape) KTEN=1 if(KTEN==1): t0=time.time() from tensorflow.keras.preprocessing import image from tensorflow.keras.optimizers import RMSprop from tensorflow import keras import tensorflow as tf import shutil from tensorflow.keras.models import Sequential from tensorflow.keras.layers import Dense from tensorflow.keras.layers import Flatten from tensorflow.keras.layers import Conv2D from tensorflow.keras.layers import MaxPooling2D from tensorflow.keras.layers import Dropout from tensorflow.keras.utils import to_categorical from sklearn.model_selection import train_test_split t1=time.time(); TEN=round(t1-t0,3); print('TEN=',TEN) if(KBU==1): buildmodel(X_train,Y_train) t2=time.time() TB=round(t2-t1,2); print('TB=',TB) model = keras.models.load_model(mod1) print('load ',mod1,' ... done') for j in range(5): if(j==0): pathT=path+'/'+LAB[0]+'/' if(j==1): pathT=path+'/'+LAB[1]+'/' if(j==2): pathT=path+'/'+LAB[2]+'/' if(j==3): pathT=path+'/'+LAB[3]+'/' if(j==4): pathT=path+'/TEST/' if(j==5): pathT=path+'/TEST2/' AITEST(pathT)