Can't use images from our site on vision api for annotation

[Andrés Tarrio]

Hi!

We're experimenting with google vision API to help us moderate images in our site. We have a different site to host our images with the domain kn3.net. However, when we try to use google vision, it refuses to load images from there.

We've tried putting the same image in another site (like imgur) and the api seems to take it fine from there.

There's no hotlinking prevention from our side.

https://developers.google.com/apis-explorer/#search/LOGO%20/vision/v1/vision.images.annotate?_h=1&resource=%257B%250A++%2522requests%2522%253A+%250A++%255B%250A++++%257B%250A++++++%2522image%2522%253A+%250A++++++%257B%250A++++++++%2522source%2522%253A+%250A++++++++%257B%250A++++++++++%2522imageUri%2522%253A+%2522http%253A%252F%252Fk62.kn3.net%252Ftaringa%252F3%252F1%252F8%252F4%252F9%252FF%252FNachone%252F5A6.jpg%2522%250A++++++++%257D%250A++++++%257D%252C%250A++++++%2522features%2522%253A+%250A++++++%255B%250A++++++++%257B%250A++++++++++%2522type%2522%253A+%2522SAFE_SEARCH_DETECTION%2522%250A++++++++%257D%250A++++++%255D%250A++++%257D%250A++%255D%250A%257D&

POST https://vision.googleapis.com/v1/images:annotate?key={YOUR_API_KEY}

{

"requests": [

 {

  "image": {

   "source": {

    "imageUri": "http://k62.kn3.net/taringa/3/1/8/4/9/F/Nachone/5A6.jpg

http://k62.kn3.net/taringa/3/1/8/4/9/F/Nachone/5A6.jpg

"

   }

  },

  "features": [

   {

    "type": "SAFE_SEARCH_DETECTION"

   }

  ]

 }

]

}

Response

{

"responses": [

 {

  "error": {

   "code": 14,

   "message": "We can not access the URL currently. Please download the content and pass it in."

  }

 }

]

}

Every time we get this. Can anyone please help? If any further information is needed I'll be happy to ask one of our devs to hop in and add detail.

cheers

[cloud-vision-discuss]

The public URL crawling generally is not meant for production use. This is intentional to prevent abusers from attacking smaller sites that they don't own because we have no knowledge of the site ownership by default.

We typically recommend our clients to either

1) download the images themselves and pass the raw bytes in

2) use Google Cloud Storage to store the files and refer to the bucket/object location

Thanks.