AI Service > Pose Estimation > API Guide

Pose Recognition API

Request

  • You can check the {appKey} and {secretKey} in the URL & Appkey menu at the top of the console.

[URI]

Method URI
POST https://pose-estimation.api.nhncloudservice.com/v1.0/appkeys/{appKey}/pose

[Request Header]

Name Value Description
Authorization {secretKey} Security key issued from the console

[Path Variable]

Name Value Description
appKey {appKey} Integrated Appkey or Service Appkey

[Request Body]

  • Input the binary data for the image file.
curl -X POST 'https://pose-estimation.api.nhncloudservice.com/v1.0/appkeys/{appKey}/pose' \
-F 'image=@sample.png' \
-H 'Authorization: ${secretKey}'

[Field]

Name Type Description
image multipart/form-data Image file

Response

[Response Body]

{
  "header": {
    "isSuccessful": true,
    "resultCode": 0,
    "resultMessage": "SUCCESS"
  },
  "result": {
    "fileType": "png",
    "numberOfPerson": 5,
    "keyPointList": [
      {
        "personId": "0",
        "inferKeypointList": [
          {
            "name": "nose",
            "part": "body",
            "x": 367.151489,
            "y": 79.552887,
            "score": 0.976
          },
          {
            "name": "left_eye",
            "part": "body",
            "x": 373.819214,
            "y": 73.5605469,
            "score": 0.985
          },
          ...
        ]
      },
      ...
      {
        "personId": 2,
        "inferKeypoints": [
          {
            "name": "nose",
            "part": "body",
            "x": 98.7361755,
            "y": 68.2939758,
            "score": 0.979
          },
          {
            "name": "left_eye",
            "part": "body",
            "x": 110.171448,
            "y": 56.0753479,
            "score": 0.983
          },
          ...
        ]
      },
    ]
  }
}

[Header]

Name Type Description
isSuccessful Boolean Recognition API success or not
resultCode Integer Result code
resultMessage String Result message (success on success, error content on failure)

[Field]

Name Type Description
fileType String File extension (.jpg, .png)
numberOfPerson Integer Number of people recognized
keyPointList[0].personId String ID assigned to distinguish recognition results
keyPointList[0].inferKeypointList[0].name String Body part name
keyPointList[0].inferKeypointList[0].part String The name of the part the body part belongs to (body, foot, face, hand)
keyPointList[0].inferKeypointList[0].x Double X coordinate of the keypoint
keyPointList[0].inferKeypointList[0].y Double Y coordinate of the keypoint
keyPointList[0].inferKeypointList[0].score Double Confidence in keypoint recognition results
TOP