---
title: "POST /videoAnalyticEvents:parse"
method: POST
path: "/videoAnalyticEvents:parse"
tags: ["Video Analytics Events"]
---

# POST /videoAnalyticEvents:parse

`POST /videoAnalyticEvents:parse`

Map a natural language query to a set of object filters to use in the deep search endpoints. The
deep search can identify a variety of objects, including people, vehicles, and some common handheld
objects such as backpacks and suitcases. The deep search can also identify the attributes of these
objects, such as the color of a person's clothing or the make of a vehicle.

## Query parameters

- `language` 'de-DE' | 'en-US' | 'es-ES' | 'fr-FR' | 'ja-JP' | 'nl-NL' | 'pl-PL' | 'pt-BR' | 'pt-PT' | 'sv-SE' | 'tr-TR' — IETF language tag used as main identifier of the language in the Eagle Eye Networks API. The format can be just the two character "primary language subtag" or a combination of the two character "primary language subtag" and 2 character "region subtag". Use `/api/v3.0/languages` API to retrieve supported language codes.

## Headers

- `Accept-Language` 'de-DE' | 'en-US' | 'es-ES' | 'fr-FR' | 'ja-JP' | 'nl-NL' | 'pl-PL' | 'pt-BR' | 'pt-PT' | 'sv-SE' | 'tr-TR' — IETF language tag used as main identifier of the language in the Eagle Eye Networks API. The format can be just the two character "primary language subtag" or a combination of the two character "primary language subtag" and 2 character "region subtag". Use `/api/v3.0/languages` API to retrieve supported language codes.

## Request body

- object
  - `query` string, required — A natural language query for searching events based on their metadata.

## Response `200`

Structured information based on the user-specified query data

- union
  - object
    - `objects__all` array[], required — List of all objects that need to be found for an event to match, each defined by a set of filters. The VMS can identify people, vehicles, and a variety of common objects. You can generate a valid filter by using the [`POST /videoAnalyticEvents:parse`](https://developer.eagleeyenetworks.com/reference/parsevideoanalytics) endpoint.
      - union[] — Set of filters the object has to match with.
        - union
          - FilterObjectClassificationV1
            - `type` 'filter.objectClassification.v1', required — Type of the filter selected
            - `class` 'person' | 'vehicle' | 'suitcase' | 'backpack' | 'handbag' | 'weapon' | 'face' | 'animal' | 'fire', required — The `class` of the object
          - FilterPersonAttributesV1 — Attributes related to a person detected in an event image
            - `type` 'filter.personAttributes.v1', required — Type of the filter selected
            - `upperBodyClothingColor` 'black' | 'blue' | 'brown' | 'gold-beige' | 'green' | 'gray' | 'silver' | 'silver-gray' | 'orange' | 'pink' | 'purple' | 'red' | 'white' | 'yellow' — Enumerations of upperbody clothing color for person.
            - `lowerBodyClothingColor` 'black' | 'blue' | 'brown' | 'gold-beige' | 'green' | 'gray' | 'silver' | 'silver-gray' | 'orange' | 'pink' | 'purple' | 'red' | 'white' | 'yellow' — Enumerations of lowerbody clothing color for person.
            - `gender` 'male' | 'female' — Enumerations of gender for person.
            - `stationary` boolean — Information regarding whether the person is moving or not
            - `helmet` boolean — Information regarding whether the person is wearing a helmet or not
            - `vest` boolean — Information regarding whether the person is wearing a vest or not
          - FilterVehicleAttributesV1 — Attributes related to the vehicle detected in an event image
            - `type` 'filter.vehicleAttributes.v1', required — Type of the filter selected
            - `color` 'beige' | 'black' | 'blue' | 'brown' | 'burgundy' | 'gold-beige' | 'green' | 'gray' | 'silver' | 'silver-gray' | 'maroon' | 'orange' | 'pink' | 'purple' | 'red' | 'teal' | 'white' | 'yellow' — Enumerations of standard colors for vehicles.
            - `make` 'abarth' | 'acura' | 'alfa_romeo' | 'aston_martin' | 'audi' | 'bentley' | 'bmw' | 'bugatti' | 'buick' | 'byd' | 'cadillac' | 'chevrolet' | 'chrysler' | 'citroen' | 'dacia' | 'daihatsu' | 'dodge' | 'dodge_ram' | 'ferrari' | 'fiat' | 'fisker' | 'ford' | 'ford_mustang' | 'garia' | 'genesis' | 'gmc' | 'harley_davidson' | 'honda' | 'hummer' | 'hyundai' | 'infiniti' | 'isuzu' | 'jaguar' | 'jeep' | 'kia' | 'lamborghini' | 'land_rover' | 'lexus' | 'lincoln' | 'lotus' | 'lucid' | 'mahindra' | 'maruti_suzuki' | 'maserati' | 'mazda' | 'mclaren' | 'mercedes_benz' | 'mini_cooper' | 'mitsubishi' | 'mg' | 'nissan' | 'oldsmobile' | 'opel' | 'perodua' | 'peugeot' | 'plymouth' | 'polaris' | 'polestar' | 'pontiac' | 'porsche' | 'proton' | 'ram' | 'renault' | 'rivian' | 'rolls_royce' | 'rover' | 'saab' | 'saturn' | 'scion' | 'skoda' | 'smart' | 'subaru' | 'suzuki' | 'tata' | 'tesla' | 'toyota' | 'vinfast' | 'volkswagen' | 'volvo' | 'wagoneer' | 'others' — Enumerations of standard makes/brands for vehicles.
            - `bodyType` 'motorbike' | 'car' | 'bus' | 'truck' | 'sedan' | 'hatchback' | 'coupe' | 'convertible' | 'minivan' | 'wagon' | 'trike' | 'roadster' | 'crossover' | 'scooter' | 'sports_car' | 'van' | 'bicycle' | 'tukTuk' | 'suv' | 'pickupTruck' | 'forklift' | 'others' — Enumerations of standard classes for vehicles. Warning: These body types are subject to future modifications
            - `parked` boolean — Information regarding whether the vehicle is parked or not
          - union — This filter can be used for the following: - To support searching for events with unique objects and eliminate the presence of duplicate results in the response - To search for all instances of a selected unique object. Searching for multiple objects in a single query will not be supported when the `objectSimilarity` filter is present in the request. When `objectId` is provided, the search resolves results using a two-tier approach: 1. **Entity match** — the objectId is resolved to a known person entity and all detections belonging to that entity are returned. The response metadata reports `type: objectSimilarity` and `resolution: entity`. 2. **Metadata fallback** — when no entity match is found and `enableFallback` is true (default), the object's metadata attributes are used for matching. The response metadata reports `type: objectSimilarity` and `resolution: metadataFallback`.
            - object
              - …
            - object
              - …
          - FilterCustomLabelsV1
            - `type` 'filter.customLabels.v1', required — Type of the filter selected
            - `label` string, required — Label user wants to search for.
          - FilterObjectAtPointV1 — Searches for all appearances of the object nearest to a clicked point in a camera frame. The caller supplies the camera, timestamp, and normalized click coordinates; the server resolves the matching entity and returns all associated detections.
            - `type` 'filter.objectAtPoint.v1', required — Type of the filter selected
            - `actor` string, required — Actor (camera) to fetch the preview frame from, in `actorType:actorId` format.
            - `timestamp` string, date-time, required — Time of the frame to inspect (RFC 3339).
            - `referencePoint` object, required — Normalized coordinates of the point clicked within the frame.
              - …
    - `queryParams` object[] — Some filters are applied as query parameters. This array gives a list of all query parameters that should be added to the request to fulfill the user request.
      - `type` string, required — The datatype of the value
      - `name` string, required — Name of the query param
      - `value` union, required — Value of the query param
        - string
        - boolean
        - integer
        - number
        - object
  - object
    - `query` string, required — Natural language text query for semantic search. Searches for persons, vehicles, and scenes based on text descriptions. Maximum 7 words recommended for optimal results. Examples: "person with red jacket", "blue sedan car", "person carrying backpack"
    - `disableSpellCheck` boolean — If true, disables automatic spell correction and uses the original query text as-is. If false or omitted, the server will attempt to correct spelling mistakes.
    - `queryParams` object[] — Some filters are applied as query parameters. This array gives a list of all query parameters that should be added to the request to fulfill the user request.
      - `type` string, required — The datatype of the value
      - `name` string, required — Name of the query param
      - `value` union, required — Value of the query param
        - string
        - boolean
        - integer
        - number
        - object

## Other responses

- `400` — The supplied object is invalid. Error detail will contain the validation error.
- `401` — You are not authenticated. Please authenticate and try again.
- `403` — You have no permission to access the specified resource.
- `404` — Referenced resource could not be found.
- `500` — Something went wrong in the server. Please try again.

---

[API](https://skmtc.net/eagleeyenetworks/apis/devices.md) · [All operations](https://skmtc.net/eagleeyenetworks/apis/devices/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/eagleeyenetworks/devices/versions/df57313de545/schema)