From 201f2fd1d87181505dafcef1ecafbd6ed2dddf3a Mon Sep 17 00:00:00 2001
From: Chunk <chunkplus@gmail.com>
Date: Sun, 19 Apr 2015 17:12:37 +0800
Subject: [PATCH] (ง •̀_•́)ง we're heading for modeling!

---
 mdata/ANALYSIS.py | 19 ++++++++++++++++++-
 mdata/crop.py     | 53 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 71 insertions(+), 1 deletion(-)
 create mode 100644 mdata/crop.py

diff --git a/mdata/ANALYSIS.py b/mdata/ANALYSIS.py
index 5b944e7..739a712 100644
--- a/mdata/ANALYSIS.py
+++ b/mdata/ANALYSIS.py
@@ -93,11 +93,28 @@ def anal_0000():
     df_ILS.hist(column='size',bins=100)
     plt.show()
 
+
+
+def pre_crop():
+    df_ILS = pd.read_csv('../res/file-tag-test.tsv', names=['hash', 'width', 'height', 'size', 'quality','chosen','class'], sep='\t')
+    print df_ILS.shape
+    print df_ILS[(df_ILS['width'] >= 300) & (df_ILS['height'] >= 300)].shape
+
+    # 300x300 4213 0.917 *
+    # 200x200 4534 0.987
+    # 400x400 932 0.202
+
+
+
+
 if __name__ == '__main__':
     # anal_ILSVRC()
     # anal_ILSVRC_Test()
-    anal_0000()
+    # anal_0000()
     # print timeit.timeit("anal_ILSVRC()", setup="from __main__ import anal_ILSVRC", number=1)
+
+
+    pre_crop()
     pass
 
 
diff --git a/mdata/crop.py b/mdata/crop.py
new file mode 100644
index 0000000..21accb8
--- /dev/null
+++ b/mdata/crop.py
@@ -0,0 +1,53 @@
+__author__ = 'chunk'
+
+import os, sys
+from PIL import Image
+from common import *
+import random
+
+base_dir = '/data/hadoop/ImageNet/ILSVRC/ILSVRC2013_DET_val/'
+category = 'Test'
+
+
+def crop_Test():
+    for path, subdirs, files in os.walk(os.path.join(base_dir, category)):
+        for name in files:
+            image = os.path.join(path, name)
+            im = Image.open(image)
+            w, h = im.size
+            if w < 300 or h < 300:
+                continue
+            left, upper = random.randint(0, w - 300), random.randint(0, h - 300)
+            im = im.crop((left, upper, left + 300, upper + 300))
+            im.save(os.path.join(base_dir, category + '_crop', name))
+
+
+
+if __name__ == '__main__':
+    timer = Timer()
+
+    timer.mark()
+    crop_Test()
+    timer.report()
+
+    pass
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
--
libgit2 0.21.2