python本機端使用訓練好Keras模型的實作

 

python本機端使用訓練好Keras模型的實作。ai機器學習辨識的程式實作要有兩個前提，一是硬體軟體，二是訓練好的模型。 

[ML python套件tensorflow-gpu keras軟體安裝說明]https://youtu.be/gEsLLIIAuGs 

[谷哥Teachable Machine訓練Keras模型操作人、貓、狗辨識]https://youtu.be/HbZ5hggMbio

Code請開：

套件匯入，觀看cuda相關部份

print('Teachable Machine之訓練好的keras model實作python code辨識部份')
!nvidia-smi
from keras.models import load_model  # TensorFlow is required for Keras to work
import cv2 
import numpy as np
import tensorflow as tf
print(tf.__version__)
import keras
print(keras.__version__)
print(tf.test.is_built_with_cuda)
tf.config.list_physical_devices('GPU')
 
print('下面上傳converted_keras.zip &欲辨識圖檔要在圖片中央')

匯入model, labels

#google colab加上files.upload
#要unzip converted_keras.zip
# Disable scientific notation for clarity
np.set_printoptions(suppress=True)
# Load the model
model = load_model("keras_model.h5", compile=False)
# Load the labels
class_names = open("labels.txt", "r").readlines()
 

測試0.jpg~5.jpg圖裁切、縮放、正則化、預測

ims=[str(i)+'.jpg' for i in range(6)]
for fn in ims:
    image = cv2.imread(fn)
    h, w, c=image.shape
    s=min(h,w)
    print(h, ',', w, ',s=',s)
#圖片裁切成正方形切
    if w<=h:
        image=image[(h-w)//2:w+(h-w)//2, :, :]
    else:
        image=image[:, (w-h)//2:h+(w-h)//2, :]
    print(image.shape)
# Resize the raw image into (224-height,224-width) pixels
    image = cv2.resize(image, (224, 224), interpolation=cv2.INTER_AREA)
#這部份在google colab要修改
    cv2.imshow('image',image)
    cv2.setWindowTitle('image', fn)
    cv2.waitKey(0)
# Make the image a numpy array and reshape it to the models input shape.
    image = np.asarray(image, dtype=np.float32).reshape(1, 224, 224, 3)
# Normalize the image array
    image = (image / 127.5) - 1
# Predicts the model
    prediction = model.predict(image)
    index = np.argmax(prediction)
    class_name = class_names[index]
    confidence_score = prediction[0][index]
    print(prediction)
# Print prediction and confidence score
    print("Class:", class_name[2:], end="")
    print("Confidence Score:", confidence_score * 100, "%")
    print('\n=================')