Merge pull request #29 from niteshgupta2711/master

Resize functionality  added

.gitignore

@@ -3,4 +3,5 @@ dataset/*
 dataset
 dist
 image_search.egg-info
-**/.vscode/*
+**/.vscode/*
+bing_image_downloader.egg-info

README.md

@@ -30,7 +30,7 @@ pip install .
 ### Usage <br />
 ```python
 from bing_image_downloader import downloader
-downloader.download(query_string, limit=100,  output_dir='dataset', adult_filter_off=True, force_replace=False, timeout=60, verbose=True)
+downloader.download(query_string, limit=100,  output_dir='dataset', adult_filter_off=True, force_replace=False, timeout=60,resize=(224,224) ,verbose=True)
 ```
 
 `query_string` : String to be searched.<br />

bing_image_downloader/bing.py

@@ -4,27 +4,49 @@
 import imghdr
 import posixpath
 import re
+from PIL import Image
+from io import BytesIO
+import io
+
 
 '''
 Python api to download image form Bing.
 Author: Guru Prasad ([email protected])
 '''
+def image_to_byte_array(image: Image) -> bytes:
+  imgByteArr = io.BytesIO()
+  image.save(imgByteArr, format="PNG")
+  imgByteArr = imgByteArr.getvalue()
+  return imgByteArr
+
+
+def resize(url,size: tuple):
 
+    response = urllib.request.urlopen(url)
+    img = Image.open(BytesIO(response.read()))
+    img=img.resize(size=size,resample=Image.LANCZOS)
+    #kl=image_to_byte_array(img)
+    # with open('pn.png','wb') as f:
+    #     f.write(kl)
+    return img
 
 class Bing:
-    def __init__(self, query, limit, output_dir, adult, timeout,  filter='', verbose=True):
+    def __init__(self, query, limit, output_dir, adult, timeout, filter='',resize=None, verbose=True):
         self.download_count = 0
         self.query = query
         self.output_dir = output_dir
         self.adult = adult
         self.filter = filter
         self.verbose = verbose
         self.seen = set()
+
 
         assert type(limit) == int, "limit must be integer"
         self.limit = limit
         assert type(timeout) == int, "timeout must be integer"
         self.timeout = timeout
+        assert (type(resize)==tuple) or (resize is None), "resize must be a tuple(height,width)"
+        self.resize=resize
 
         # self.headers = {'User-Agent': 'Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:60.0) Gecko/20100101 Firefox/60.0'}
         self.page_counter = 0
@@ -54,16 +76,30 @@ def get_filter(self, shorthand):
 
 
     def save_image(self, link, file_path):
-        request = urllib.request.Request(link, None, self.headers)
-        image = urllib.request.urlopen(request, timeout=self.timeout).read()
-        if not imghdr.what(None, image):
-            print('[Error]Invalid image, not saving {}\n'.format(link))
-            raise ValueError('Invalid image, not saving {}\n'.format(link))
-        with open(str(file_path), 'wb') as f:
-            f.write(image)
+        if not self.resize:
 
-
+            request = urllib.request.Request(link, None, self.headers)
+            image = urllib.request.urlopen(request, timeout=self.timeout).read()
+            if not imghdr.what(None, image):
+                print('[Error]Invalid image, not saving {}\n'.format(link))
+                raise ValueError('Invalid image, not saving {}\n'.format(link))
+            with open(str(file_path), 'wb') as f:
+                f.write(image)
+        elif self.resize:
+            request = urllib.request.Request(link, None, self.headers)
+
+            img=resize(request,size=self.resize)
+            image=image_to_byte_array(img)
+            # if not imghdr.what(None, image):
+            #     print('[Error]Invalid image, not saving {}\n'.format(link))
+            #     raise ValueError('Invalid image, not saving {}\n'.format(link))
+            with open(str(file_path), 'wb') as f:
+                f.write(image)
+
+
+
     def download_image(self, link):
+
         self.download_count += 1
         # Get the image link
         try:

bing_image_downloader/downloader.py

@@ -9,7 +9,7 @@
 
 
 def download(query, limit=100, output_dir='dataset', adult_filter_off=True, 
-force_replace=False, timeout=60, filter="", verbose=True):
+force_replace=False, timeout=60, filter="",resize=None, verbose=True):
 
     # engine = 'bing'
     if adult_filter_off:
@@ -34,7 +34,7 @@ def download(query, limit=100, output_dir='dataset', adult_filter_off=True,
         sys.exit(1)
 
     print("[%] Downloading Images to {}".format(str(image_dir.absolute())))
-    bing = Bing(query, limit, image_dir, adult, timeout, filter, verbose)
+    bing = Bing(query, limit, image_dir, adult, timeout, filter,resize, verbose)
     bing.run()