printf("ho_tari\n");
손글씨 인식 본문
<randomforestclassifier.py>
from sklearn.datasets import fetch_openml
mnist = fetch_openml('mnist_784')
print(mnist.data)
import matplotlib.pyplot as plt
plt.imshow(mnist.data.values[0].reshape(28,28),cmap = 'gray')
from sklearn.model_selection import train_test_split
X_train,X_test,y_train,y_test = train_test_split(mnist.data,mnist.target,test_size=0.1)
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score
clf = RandomForestClassifier()
clf.fit(X_train,y_train)
pred = clf.predict(X_test)
print(accuracy_score(y_test,pred))
import glob
from PIL import Image
import numpy as np
import warnings
warnings.filterwarnings('ignore')
for path in glob.glob('img/*.png'):
#print(path)
img = Image.open(path).convert('L')
#print(img)
#plt.imshow(img,cmap='gray')
data = np.resize(img,(28,28))
data = 255-data
plt.imshow(data,cmap='gray')
data = data.reshape(1,-1)
pred = clf.predict(data)
print(pred)
plt.show()
'Python' 카테고리의 다른 글
SNS-project(1) (client UI 제작) (0) | 2023.11.02 |
---|---|
titanic_csv 평가지표 (0) | 2023.10.31 |
label encoding (0) | 2023.10.30 |
평가지표 (0) | 2023.10.30 |
pandas (0) | 2023.10.30 |