From 6f9dfb9d4c4257861c3354267a4701c3354370f0 Mon Sep 17 00:00:00 2001
From: zhai_pro <xin_soft@foxmail.com>
Date: Sun, 27 Jan 2019 17:30:49 +0800
Subject: [PATCH] =?UTF-8?q?=E7=9C=8B=E7=9C=8B=E6=B7=B1=E5=BA=A6=E5=AD=A6?=
 =?UTF-8?q?=E4=B9=A0=E5=90=8E=E7=9A=84=E6=A1=86=E6=9E=B6=E6=98=AF=E5=90=A6?=
 =?UTF-8?q?=E8=83=BD=E8=AF=86=E5=88=AB=E5=89=8D=E6=89=80=E6=9C=AA=E8=A7=81?=
 =?UTF-8?q?=E7=9A=84=E5=9B=BE=E7=89=87?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 mlearn_for_image.py | 20 ++++++++++++++++++--
 1 file changed, 18 insertions(+), 2 deletions(-)

diff --git a/mlearn_for_image.py b/mlearn_for_image.py
index 2fb483b..1b7b8f0 100644
--- a/mlearn_for_image.py
+++ b/mlearn_for_image.py
@@ -1,4 +1,6 @@
 # coding: utf-8
+import sys
+
 import cv2
 import numpy as np
 from keras import models
@@ -31,7 +33,7 @@ def load_data():
     return (train_x, train_y, sample_weight), (test_x, test_y)
 
 
-def main():
+def learn():
     (train_x, train_y, sample_weight), (test_x, test_y) = load_data()
     datagen = ImageDataGenerator(horizontal_flip=True,
                                  vertical_flip=True)
@@ -62,5 +64,19 @@ def main():
     model.save('12306.image.model.h5', include_optimizer=False)
 
 
+def predict(fn):
+    imgs = cv2.imread(fn)
+    imgs = cv2.resize(imgs, (67, 67))
+    imgs = imgs / 255.0
+    imgs.shape = (-1, 67, 67, 3)
+    model = models.load_model('12306.image.model.h5')
+    labels = model.predict(imgs)
+    print(labels.max(axis=1))
+    print(labels.argmax(axis=1))
+
+
 if __name__ == '__main__':
-    main()
+    if len(sys.argv) >= 2:
+        predict(sys.argv[1])
+    else:
+        learn()