Teachable Machine

Teachable Machine Access

Part 01. Open in Web

Part 02. Open in Python

Advance Preparation

Running the virtual environment

• 'Windows 키 + R'을 눌러 관리자 권한으로 명령 프롬프트를 실행한다.

• 명령 프롬프트에서 다음 명령어를 실행하여 '새로운 가상환경'을 생성한다.

C:\Project\Keras> conda create –n venvtm python=3.8

• 다음 명령으로 '가상환경'을 실행한다.

C:\Project\Keras> activate venvtm

• 다음 명령으로 가상환경에 주피터 노트북을 설치한다.

C:\Project\Keras> conda install –n venvtm ipython notebook

Library Installation

• python의 pip를 이용하여 'TensorFlow', 'Pillow', 'Numpy', 'OpenCV', 'pySerial'을 설치한다.

pip install tensorflow==2.3.1 (Python 3.8.6 버전에서 설치)

pip install pillow

pip install numpy

pip install opencv-python 또는 pip install opencv-contrib-python

pip install pyserial

Jupyter Notebook Execute

• 다음 명령으로 'Jupyter Notebook'을 실행한다.

C:\Project\Keras> jupyter notebook

Prediction Picture

Python Code (Prediction Picture)

import tensorflow.keras

from PIL import Image, ImageOps

import numpy as np

# Disable scientific notation for clarity

np.set_printoptions(suppress=True)

# Load the model

model = tensorflow.keras.models.load_model('keras_model.h5', compile=False)

# Create the array of the right shape to feed into the keras model

# The 'length' or number of images you can put into the array is determined

# by the first position in the shape tuple, in this case 1.

data = np.ndarray(shape=(1, 224, 224, 3), dtype=np.float32)

# Replace this with the path to your image

image = Image.open('test_photo.jpg')

# resize the image to a 224x224 with the same strategy as in TM2:

# resizing the image to be at least 224x224 and then cropping from the center

size = (224, 224)

image = ImageOps.fit(image, size, Image.ANTIALIAS)

# turn the image into a numpy array

image_array = np.asarray(image)

# display the resized image

image.show()

# 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 = model.predict(data)

print(prediction)

Run (Prediction Picture)

• 'Shift+Enter' 키를 눌러 실행하면 아래와 같은 예측 결과를 얻는다.

[[0.999841 0.00015903]]

Python Code (Prediction WebCAM)

import tensorflow.keras

from PIL import Image, ImageOps

import numpy as np

import cv2

np.set_printoptions(suppress=True)

model = tensorflow.keras.models.load_model('keras_model.h5', compile=False)

data = np.ndarray(shape=(1, 224, 224, 3), dtype=np.float32)

cap = cv2.VideoCapture(0)

while(True):

ret, frame = cap.read()

cv2.imshow('frame', frame)

image = cv2.resize(frame, dsize=(224, 224), interpolation=cv2.INTER_CUBIC)

image_array = np.asarray(image)

normalized_image_array = (image_array.astype(np.float32) / 127.0) - 1

data[0] = normalized_image_array

prediction = model.predict(data)

print(prediction)

if prediction[:, 0] > 0.7 :

send = (str('a')+'\n').encode("utf-8")

print(send)

if prediction[:, 1] > 0.7 :

send = (str('b')+'\n').encode("utf-8")

print(send)

if cv2.waitKey(20) & 0xFF == ord('q'):

break

cap.release()

cv2.destroyAllWindows()

Python Code (Prediction WebCAM & Serial Communication)

import tensorflow.keras

from PIL import Image, ImageOps

import numpy as np

import cv2

import serial

ser = serial.Serial(

# port='/dev/cu.usbmodem141201',

port='COM3',

baudrate=9600,

)

np.set_printoptions(suppress=True)

model = tensorflow.keras.models.load_model('keras_model.h5', compile=False)

data = np.ndarray(shape=(1, 224, 224, 3), dtype=np.float32)

cap = cv2.VideoCapture(0)

while(True):

ret, frame = cap.read()

cv2.imshow('frame',frame)

image = cv2.resize(frame, dsize=(224, 224), interpolation=cv2.INTER_CUBIC)

image_array = np.asarray(image)

normalized_image_array = (image_array.astype(np.float32) / 127.0) - 1

data[0] = normalized_image_array

prediction = model.predict(data)

print(prediction)

# print(prediction[:, 0])

# print(prediction[:, 1])

# print(prediction[:, 2])

if prediction[:, 0] > 0.7 :

send = (str('a')+'\n').encode("utf-8")

ser.write(send)

print(send)

if prediction[:, 1] > 0.7 :

send = (str('b')+'\n').encode("utf-8")

ser.write(send)

print(send)

if cv2.waitKey(20) & 0xFF == ord('q'):

break

cap.release()

cv2.destroyAllWindows()