[fix] handling empty resultset II.

searx/engines/google_images.py

     search_res = loads(resp.text)
     if 'responseData' not in search_res:
         return []
-    if 'results' not in search_res['responseData']:
+    if 'results' not in search_res['responseData'] and search_res['responseData']['results']:
         return []
     for result in search_res['responseData']['results']:
         url = result['originalContextUrl']