%matplotlib inline
import numpy as np
import os
import matplotlib.pyplot as plt
fnames = np.array([f'train/{f}' for f in sorted(os.listdir('train/'))])
labels = np.array([(0 if 'cat' in fname else 1) for fname in fnames])

img = plt.imread(fnames[0])
#plt.imshow(img)
print(fnames)

from keras.applications.resnet50 import ResNet50
ResNet50_model = ResNet50(weights='imagenet')
 

from sklearn.model_selection import train_test_split
X_train,X_valid, y_train, y_valid =train_test_split(fnames,labels,test_size=0.2, random_state=1)

from keras.preprocessing import image                  
from tqdm import tqdm

def path_to_tensor(img_path):
    # 用PIL載入RGB影象為PIL.Image.Image型別
    img = image.load_img(img_path, target_size=(224, 224))
    # 將PIL.Image.Image型別轉化為格式為(224, 224, 3)的3維張量
    x = image.img_to_array(img)
    # 將3維張量轉化為格式為(1, 224, 224, 3)的4維張量並返回
    return np.expand_dims(x, axis=0)

def paths_to_tensor(img_paths):
    list_of_tensors = [path_to_tensor(img_path) for img_path in tqdm(img_paths)]
    return np.vstack(list_of_tensors)
 

from keras.applications.resnet50 import preprocess_input, decode_predictions
def ResNet50_predict_labels(img_path):
    # 返回img_path路徑的影象的預測向量
    img = preprocess_input(path_to_tensor(img_path))
    return np.argmax(ResNet50_model.predict(img))

def detector(img_path):
    prediction = ResNet50_predict_labels(img_path)
    if (prediction <= 268) & (prediction >= 151):
        index=1
    elif (prediction <= 285) & (prediction >= 281):
        index=0
    else:
        index=prediction
    return index 
 

for i in range(100):
    img = plt.imread(X_train[i])
    plt.imshow(img);
    plt.show() 
    index_num=detector(X_train[i])
    if index_num ==0:
        call='cat'
    elif index_num ==1:
        call='dog'
    else:
        call=index_num
    print(f'the picture is {call}\n')
    print(index_num)