Create cnn_class.py

Convolutional Neural Networks/Cat and dog/cnn_class.py

@@ -0,0 +1,32 @@
+import numpy as np
+import matplotlib.pyplot as plt
+import matplotlib.image as mpimg
+from keras.models import load_model
+from keras.preprocessing import image
+
+
+# dimensions of our images    -----   are these then grayscale (black and white)?
+img_width, img_height = 64, 64
+
+# load the model we saved
+model = load_model('dog_cat_model.h5')
+
+#Image source
+img_src = 'dataset/training_set/cats/cat.100.jpg'
+
+# Get test image ready
+img = mpimg.imread(img_src)
+test_image_main = image.load_img(img_src, target_size=(img_width, img_height))
+test_image = image.img_to_array(test_image_main)
+test_image = np.expand_dims(test_image, axis=0)
+
+result = model.predict(test_image, batch_size=1)
+
+if result[0][0]==1.0:
+  answer = 'Dog'
+else:
+  answer = 'cat'
+
+plt.imshow(img)
+plt.title(answer)
+plt.show()