ics

The Image Classification Service (ICS) allows users to submit image URLs and get the linked image classified (tagged) based on their perceived content. To do this, the service relies on the image recognition service Ximilar.

The project was developed within the VMware Bulgaria Talent Boost 2023.

Table of content:

• Overall architecture
• Requests
• Demo and snaps

Overall architecture

Requests:

• GET /images -> lists all analysed images from the DB

Parameter	Type	Required	Description
labels	Array of strings	No	Specifies the labels to filter the images. If provided, images with matching labels will be returned.

Example response:

[
    {
        "imageId": 9,
        "imageUrl": "https://example.com/image.jpg",
        "createdAt": "2023-05-12T23:25:26.762859",
        "updatedAt": "2023-05-12T23:25:26.762859",
        "width": 1485.0,
        "height": 835.0,
        "imageService": "Ximilar",
        "labels": [
            {
                "labelId": 67,
                "labelDescription": "example"
            },
            {
                "labelId": 65,
                "labelDescription": "label"
            }]
      },
      {
        "imageId": 10,
        "imageUrl": "https://example.com/image2.jpg",
        "createdAt": "2023-06-12T23:25:26.762859",
        "updatedAt": "2023-06-12T23:25:26.762859",
        "width": 1485.0,
        "height": 835.0,
        "imageService": "Ximilar",
        "labels": [
            {
                "labelId": 61,
                "labelDescription": "example2"
            },
            {
                "labelId": 65,
                "labelDescription": "label"
            }]
      }
]

• POST /images

Parameter	Type	Required	Description
noCache	boolean	No	Defaults to false. When true, the service shall always perform categorisation, replacing any prior tags for the same image

Example body:

{
 "records": [
     {
         "_url": "https://example.com/image.jpg"
     }
 ]
}

Example response:

 {
    "imageId": 10,
    "imageUrl": "https://example.com/image.jpg",
    "createdAt": "2023-05-13T12:25:58.22273",
    "updatedAt": "2023-05-13T12:25:58.22273",
    "width": 1338.0,
    "height": 896.0,
    "imageService": "Ximilar",
    "labels": [
        {
            "labelId": 68,
            "labelDescription": "easter"
        },
        {
            "labelId": 3,
            "labelDescription": "isolated"
        },
        {
            "labelId": 4,
            "labelDescription": "ladybug"
        }
      ]
   }

• GET /images/{id} -> returns an image with a specific image id
• DELETE /images/{id} -> deletes image with a specific imageID from the DB

---

• GET /labels - lists all labels from the DB Example response:

  [
    {
        "labelId": 1,
        "labelDescription": "illustration"
    },
    {
        "labelId": 2,
        "labelDescription": "vector"
    },
    {
        "labelId": 3,
        "labelDescription": "isolated"
    },
  ]

• POST /labels Example body:

{
    "name":"newLabel"
}

Example response:

   {
    "labelId": 269,
    "labelDescription": "newLabel"
}

• GET /labels/{id} -> returns a label with a specific label id
• DELETE /labels/{id} -> deletes a label with a specific label id from the DB

Demo and snaps

Video:

demo.mp4