Bing画像検索API Pythonクライアント

学習画像を収集するための、Bing画像API Pythonクライアント。
保存ディレクトリとか検索ワードの指定をパラメータで引き渡すようにして少し使いやすくしたつもり。収集する画像の最大枚数を指定することも可能。

	#conding:utf-8
	import sys
	import os
	import urllib
	import urllib2
	import json
	import requests
	KEY = '<Your Bing Developer Key>'
	OUTPUT = '/images/bing/'
	MAX = 100
	count = 1
	def bing_search(query, directory, skip):
	global count
	bing_url = 'https://api.datamarket.azure.com/Bing/Search/v1/Image'
	print 'search count: ' + str(count) + ', url: ' + bing_url + ', skip: ' + str(skip)
	pm = urllib2.HTTPPasswordMgrWithDefaultRealm()
	pm.add_password(None, bing_url, KEY, KEY)
	handler = urllib2.HTTPBasicAuthHandler(pm)
	opener = urllib2.build_opener(handler)
	urllib2.install_opener(opener)
	if skip > 0:
	params = urllib.urlencode({'Query': "'" + query + "'", 'Adult': "'Off'", '$skip': skip ,'$format': 'json'})
	else:
	params = urllib.urlencode({'Query': "'" + query + "'", 'Adult': "'Off'", '$format': 'json'})
	response = urllib2.urlopen(bing_url+'?'+params)
	data = json.loads(response.read())
	results = data['d']['results']
	for item in results:
	if count > MAX:
	print 'finish. count: ' + str(MAX)
	return
	image_url = item['MediaUrl']
	root,ext = os.path.splitext(image_url)
	if ext.lower() == '.jpg':
	print image_url,
	try:
	r = requests.get(image_url)
	fname = OUTPUT + directory + "/bing%04d.jpg" % count
	f = open(fname, 'wb')
	f.write(r.content)
	f.close()
	print "...save", fname
	except:
	print "error", fname
	count += 1
	bing_search(query, directory, count)
	if __name__ == '__main__':
	argvs = sys.argv
	argc = len(argvs)
	if(argc != 3):
	print 'Usage: python %s query directory' % argvs[0]
	quit()
	query = argvs[1]
	directory = argvs[2]
	print 'get bing image: %s ' % query
	bing_search(query, directory, 0)

view raw bing_imagegetter.py hosted with ❤ by GitHub