Imagenet迁移学习图片分类示例

import keras
from keras.applications import Xception
from keras.applications.vgg19 import VGG19

model_xception = Xception(weights="imagenet")
model_vgg19 = VGG19(weights='imagenet')

from keras.preprocessing.image import load_img,img_to_array
import matplotlib.pyplot as plt
import numpy as np
%matplotlib inline

image = load_img("./ship.jpeg",target_size=(224,224))
plt.imshow(image)

image = img_to_array(image)
image = np.expand_dims(image,axis = 0)
image.shape

import json
from keras.applications import imagenet_utils

preds = model_xception.predict(image)
# P = imagenet_utils.decode_predictions(preds) #去下载imagenet对应的类别表json
with open('./imagenet_class_index.json') as f:  #也可以直接下载好 进行引用
    js = json.load(f)

for p in preds: #遍历每个input
    s= [(js[str(i)],p[i]) for i in p.argsort()[-1:-4:-1]] #输出最大3个
    print(s)