目標: 用瀏覽器上傳圖案到Web後端進行預測, 並將模型預測結果回傳至瀏覽器。
#teachable machine #Flask #Keras
#Install required packages
pip install flask==2.1.1
pip install Pillow==9.1.0
pip install tensorflow==2.3.1
pip install keras==2.4.3
專案目錄結構如下, 其中model 資料匣 要放置AI 模型檔及Label 文字檔。凡應用是上傳圖片，由模型預測分類結果的, 都可以套用本專案程式架構。在不想改程式碼, 就直接模型檔名改成keras_model.h5 ，類別的文字檔改成 labels.txt。
執行主程式 main.py 後, 會載入模型同時啓動Web Server (Flask)
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| from keras.models import load_model | |
| from PIL import Image, ImageOps #Install pillow instead of PIL | |
| import numpy as np | |
| import os | |
| # Disable scientific notation for clarity | |
| np.set_printoptions(suppress=True) | |
| class AI_model(): | |
| def __init__(self,model,class_name): | |
| # Load the model | |
| self.model = load_model(model, compile=False) | |
| # Load the labels | |
| self.class_names = open(class_name, 'r').readlines() | |
| def predict(self,img_file): | |
| image=self.get_input_image(img_file) | |
| data = np.ndarray(shape=(1, 224, 224, 3), dtype=np.float32) | |
| #turn the image into a numpy array | |
| image_array = np.asarray(image) | |
| # Normalize the image | |
| normalized_image_array = (image_array.astype(np.float32) / 127.0) - 1 | |
| # Load the image into the array | |
| data[0] = normalized_image_array | |
| # run the inference | |
| prediction = self.model.predict(data) | |
| index = np.argmax(prediction) | |
| class_name = self.class_names[index] | |
| print(self.class_names) | |
| confidence_score = prediction[0][index] | |
| return confidence_score,class_name | |
| def get_input_image(self,img_file): | |
| image = Image.open(img_file).convert('RGB') | |
| size = (224, 224) | |
| image = ImageOps.fit(image, size, Image.Resampling.LANCZOS) | |
| return image | |
| if __name__ == '__main__': | |
| model_dir='model' | |
| model_file='keras_Model.h5' | |
| label_file='labels.txt' | |
| upload_img='下載.png' | |
| model_file=os.path.join(model_dir,model_file) | |
| class_file=os.path.join(model_dir,label_file) | |
| model=AI_model(model_file,class_file) | |
| img_file=os.path.join('upload',upload_img) | |
| conf,label=model.predict(img_file) | |
| print('Class:', label, end='') | |
| print('Confidence score:', conf) | |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| from aimodel import AI_model #joseph | |
| @app.route('/success', methods = ['POST']) | |
| def success(): | |
| if request.method == 'POST': | |
| f = request.files['file'] | |
| f.save('upload/'+f.filename) | |
| #predict the image | |
| img_file=os.path.join('upload',f.filename) | |
| prob,label=model.predict(img_file) | |
| print('Class:', label, end='') | |
| print('Confidence score:', prob) | |
| return render_template("success.html", name = img_file,class_name=label,confidence=prob) | |
| if __name__ == '__main__': | |
| init_gpu() | |
| model_dir='model' | |
| model_file='keras_Model.h5' | |
| label_file='labels.txt' | |
| model_file=os.path.join(model_dir,model_file) | |
| class_file=os.path.join(model_dir,label_file) | |
| model=AI_model(model_file,class_file) | |
| #run flask web engine | |
| app.run(host= '0.0.0.0', port=3100 ,debug = True) |
, of a random variable X is weighted average of the possible values of X, weight by their corresponding probabilities: