İrem Documentation

Object Recognition API

This API takes a picture and recognizes the objects in that picture with the correction percentage and returns an Json abject with that data. This API can be used in our project to help the users learn the vocabulary of the object in that picture.

Usage

• Url for API: /obj/
• Returns a Json object such that:

---

{
  "images": [
    {
      "classifiers": [
        {
          "classifier_id": "default",
          "name": "default",
          "classes": [
            {
              "class": "<Object Name>",
              "score": <Score of the object>,
              "type_hierarchy": "<Type of the object>"
            }
          ]
        }
      ],
      "source_url": "<Url of the image>",
      "resolved_url": "<Url of the image>"
    }
  ],
  "images_processed": 1,
  "custom_classes": 0
} 

---

Class means basically the name of the object in the picture, score is how confident the object detection is about the classification. Also the api doesn't return any classification that has a score that is less than 0.5. Type hierarych is the bigger classes that the class of the object belongs to. For example a dog is an animal too. Custom classes papameter will always be 0 for our api as we use the classes that are built-in in the watson-object-detection api which is the api that I am using for this case. Images processed will be one as we are giving single image to the api. Source url and resolved will be the same as well.

Example

• Picture:

• Url Input: /obj/https://watson-developer-cloud.github.io/doc-tutorial-downloads/visual-recognition/640px-IBM_VGA_90X8941_on_PS55.jpg

• Json Output:

---

{
  "images": [
    {
      "classifiers": [
        {
          "classifier_id": "default",
          "name": "default",
          "classes": [
            {
              "class": "circuit board",
              "score": 0.578,
              "type_hierarchy": "/electrical device/computer circuit/circuit board"
            },
            {
              "class": "computer circuit",
              "score": 0.755
            },
            {
              "class": "electrical device",
              "score": 0.757
            },
            {
              "class": "disk controller",
              "score": 0.553,
              "type_hierarchy": "/controller/disk controller"
            },
            {
              "class": "controller",
              "score": 0.558
            },
            {
              "class": "central processing unit",
              "score": 0.535
            },
            {
              "class": "PC board",
              "score": 0.501,
              "type_hierarchy": "/electrical device/computer circuit/PC board"
            },
            {
              "class": "CPU board",
              "score": 0.5,
              "type_hierarchy": "/electrical device/computer circuit/CPU board"
            },
            {
              "class": "electronic equipment",
              "score": 0.6
            },
            {
              "class": "memory device",
              "score": 0.599
            },
            {
              "class": "microchip",
              "score": 0.592
            },
            {
              "class": "jade green color",
              "score": 0.838
            },
            {
              "class": "emerald color",
              "score": 0.787
            }
          ]
        }
      ],
      "source_url": "https://watson-developer-cloud.github.io/doc-tutorial-downloads/visual-recognition/640px-IBM_VGA_90X8941_on_PS55.jpg",
      "resolved_url": "https://watson-developer-cloud.github.io/doc-tutorial-downloads/visual-recognition/640px-IBM_VGA_90X8941_on_PS55.jpg"
    }
  ],
  "images_processed": 1,
  "custom_classes": 0
}

---